&EFW
EPA 600/R-10/076F | February 2013 | www.epa.gov/ord
United States
Environmental Protection
Agency
Integrated Science Assessment
for Ozone and Related
Photochemical Oxidants
Office of Research and Development
National Center for Environmental Assessment-RTP Division
U.S. Environmental Protection Agency
Research Triangle Park, NC
-------
United States
Environmental Protection February 2013
A9ency EPA/600/R-10/076F
Integrated Science Assessment for Ozone
and Related Photochemical Oxidants
National Center for Environmental Assessment-RTP Division
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
-------
DISCLAIMER
This document has been reviewed in accordance with U.S. Environmental Protection
Agency policy and approved for publication. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
-------
TABLE OF CONTENTS
OZONE PROJECT TEAM xxiv
AUTHORS, CONTRIBUTORS, AND REVIEWERS xxvii
CLEAN AIR SCIENTIFIC ADVISORY COMMITTEE OZONE NAAQS REVIEW PANEL xxxiii
ACRONYMS AND ABBREVIATIONS xxxv
PREAMBLE li
Process of ISA Development II
Figure I Illustration of the key steps in the process of the review of National
Ambient Air Quality Standards.
Figure II Illustration of processes for literature search and study selection used for
development of ISAs.
Figure III Characterization of the general process of ISA development. Ivii
EPA Framework for Causal Determination Iviii
Evaluating Evidence for Inferring Causation Iviii
Consideration of Evidence from Scientific Disciplines lix
Application of Framework for Causal Determination Ixiv
Table I Aspects to aid in judging causality. Ixv
Determination of Causality Ixvi
Table II Weight of evidence for causal determination. Ixviii
Quantitative Relationships: Effects on Human Populations Ixix
Quantitative Relationships: Effects on Ecosystems or Public Welfare Ixx
Concepts in Evaluating Adversity of Health Effects Ixx
Concepts in Evaluating Adversity of Ecological Effects Ixxi
References Ixxiii
LEGISLATIVE AND HISTORICAL BACKGROUND Ixxv
Legislative Requirements for the NAAQS Review Ixxv
History of the NAAQS for Ozone Ixxvi
Table III Summary of primary and secondary NAAQS promulgated for O3 during
the period 1971-2008. Ixxvii
References Ixxxi
1 EXECUTIVE SUMMARY 1-1
Introduction and Purpose 1-1
Scope and Methods 1-1
Ambient Ozone Concentrations 1-2
Human Exposure to Ozone 1-3
Dosimetry and Modes of Action 1-4
Integration of Ozone Health Effects 1-4
Table 1 -1 Summary of O3 causal determinations by exposure duration and health
outcome. 1-5
Respiratory Effects 1-6
Mortality Effects 1-7
Cardiovascular Effects 1-7
Populations Potentially at Increased Risk 1-8
Integration of Effects on Vegetation and Ecosystems 1-8
Table 1 -2 Summary of O3 causal determination for welfare effects. 1 -9
111
-------
Visible Foliar Injury
Growth, Productivity, Carbon Storage and Agriculture
Water Cycling
Below Ground Processes
Community Composition
Air Quality Indices and Exposure-Response
The Role of Tropospheric Ozone in Climate Change and UV-B Shielding Effects
Radiative Forcing and Climate Change
UV-B Shielding Effects
Table 1 -3 Summary of O3 causal determination for climate change and UV-B
shielding effects.
Conclusion
1-10
1-10
1-11
1-11
1-11
1-12
1-12
1-13
1-13
1-14
1-14
2 INTEGRATIVE SUMMARY 2-1
2.1 ISA Development and Scope 2-1
2.2 Atmospheric Chemistry and Ambient Concentrations 2-4
2.2.1 Physical and Chemical Processes 2-4
2.2.2 Background O3 Concentrations 2-5
Figure 2-1 Mean daily average maximum 8-h avg O3 concentrations in surface air,
for spring and summer 2006. 2-7
2.2.3 Monitoring 2-8
2.2.4 Ambient Concentrations 2-8
2.3 Human Exposure 2-10
2.4 Dosimetry and Mode of Action 2-12
2.5 Integration of Ozone Health Effects 2-14
2.5.1 Conclusions from Previous O3 AQCDs 2-15
2.5.2 Summary of Causal Determinations 2-16
Table 2-1 Summary of evidence from epidemiologic, controlled human exposure,
and animal toxicological studies on the health effects associated with
short- and long-term exposure to O3. 2-20
2.5.3 Integrated Synthesis of Evidence for Health Effects 2-24
2.5.4 Policy Relevant Considerations 2-29
2.6 Integration of Effects on Vegetation and Ecosystems 2-35
2.6.1 Visible Foliar Injury 2-35
Figure 2-2 An illustrative diagram of the major endpoints that O3 may affect in plants
and ecosystems. 2-36
Table 2-2 Summary of O3 causal determinations for vegetation and ecosystem
effects. 2-37
2.6.2 Growth, Productivity, Carbon Storage and Agriculture 2-38
2.6.3 Water Cycling 2-40
2.6.4 Below-ground Processes 2-41
2.6.5 Community Composition 2-42
2.6.6 Policy Relevant Considerations 2-43
2.7 The Role of Tropospheric O3 in Climate Change and UV-B Shielding Effects 2-45
2.7.1 Tropospheric Ozone as a Greenhouse Gas 2-45
Figure 2-3 Schematic illustrating the effects of tropospheric O3 on climate; including
the relationship between precursor emissions, tropospheric O3
abundance, radiative forcing, climate response, and climate impacts. 2-46
2.7.2 Tropospheric Ozone and UV-B Shielding Effects 2-47
2.8 Summary of Causal Determinations for Health Effects and Welfare Effects 2-48
Table 2-3 Summary of O3 causal determinations by exposure duration and health
outcome. 2-49
Table 2-4 Summary of O3 causal determination for welfare effects. 2-50
Table 2-5 Summary of O3 causal determination for climate and UV-B shielding
effects. 2-51
References 2-52
IV
-------
3 ATMOSPHERIC CHEMISTRY AND AMBIENT CONCENTRATIONS
3-1
3.1 Introduction
3.2 Physical and Chemical Processes
Figure 3-1 Schematic overview of photochemical processes influencing stratospheric
3.2.1
3.2.2
3.2.3
3.2.4
and tropospheric O3.
Sources of Precursors Involved in O3 Formation
Figure 3-2 Estimated anthropogenic emissions of O3 precursors for 2005.
Gas Phase Reactions Leading to O3 Formation
Multiphase Processes
Temperature and Chemical Precursor Relationships
Figure 3-3 Measured concentrations of O3 and NOZ._
3.3 Atmospheric Modeling
Figure 3-4
Figure 3-5
3.3.1 Global Scale CTMs
Figure 3-6
Sample Community Multi-scale Air Quality (CMAQ) modeling domains.
Main components of a comprehensive atmospheric chemistry modeling
system, such as the U.S. EPA's Community Multi-scale Air Quality
(CMAQ) modeling system.
3.4
Comparison of global chemical-transport model (CTM) predictions of daily
maximum 8-h avg O3 concentrations and multi-model mean with monthly
averaged CASTNET observations in the Intermountain West and
Southeast Regions of the U.S.
Background O3 Concentrations
Figure 3-7 Schematic overview of contributions to North American (NA) background
concentrations of O3.
3.4.1 Contributions from Natural Sources
3.4.2 Contributions from Anthropogenic Emissions
Figure 3-8 Time series of MDA8 O3 concentrations (ppm) measured at Trinidad
3.4.3
Head, CA, from April 18, 2002 through December 31, 2009.
Estimating Background Concentrations
Figure 3-9
Figure 3-10
Figure 3-11
Figure 3-12
Figure 3-13
Mean MDA8 O3 concentrations in surface air for spring and summer
2006 calculated by GEOS-Chem for the base case (Base),
U.S. background (USB), and NA background (NAB).
Spring and summer mean Canadian and Mexican (CM) contributions to
MDA8 O3 determined as the difference between the U.S. background
and NA background.
MDA8 O3 concentrations for spring (March-May) and summer (June-
August) 2006 simulated by GEOS-Chem vs. measured by the ensemble
of CASTNET sites in the Intermountain West, Northeast, Great Lakes,
and Southeast.
Frequency distributions of MDA8 O3 concentrations in March- August
2006 for the ensemble of low-altitude (<1,500 meters) and high-altitude
CASTNET sites (>1,500 meters) in the U.S.
Mean MDA8 O3 concentrations in surface air during spring and summer
_ 3-1
_ 3-1
_3-3
_3-5
_3-6
_3-10
_3-14
_3-17
_3-21
3-22
3-23
_3-24
3-27
_3-29
3-30
_3-32
_3-32
_3-36
_3-38
_3-40
_3-42
3-44
3-45
3-48
2006 (top) calculated by GEOS-Chem/CAMx for the base case (Base,
top) and NA background (NAB, bottom). 3-50
Figure 3-14 Monthly average MDA8 O3 concentrations observed (Obs) and predicted
for the base case and NA background (NAB) by GEOS-Chem (GC) and
GEOS-Chem/CAMx (CX) at CASTNET sites above 1,500 meters
elevation (upper panel) and CASTNET sites below 1,500 meters
elevation (lower panel). 3-51
Figure 3-15 Annual 4th-highest MDA8 O3 predicted by GEOS-Chem (0.5° x 0.667°)
for the base case (Base) with corresponding U.S. background (USB) and
NA background (NAB) MDA8 O3 for the same days in 2006. 3-56
Figure 3-16 Annual 4th-highest MDA8 O3 predicted by CAMx for the base case
(Base) and corresponding NA background (NAB) MDA8 O3 for the same
days in 2006. 3-57
Table 3-1 Comparison of Zhang et al. (2011) and Emery et al. (2012) results for
MDA8 O3 concentrations (ppbv) with measurements at selected
CASTNET sites. 3-60
Table 3-2 Comparison of annual 4th-highest MDA8 O3 concentrations measured at
CASTNET sites in 2006 with MDA8 O3 concentrations simulated by the
GEOS-Chem and CAMx base case models. 3-61
-------
3.5 Monitc
3.5.1
3.5.2
3.5.3
3.5.4
3.5.5
3.5.6
3.6 Ambie
3.6.1
iring
Routine Monitoring Technigues
Precision and Bias
Table 3-3 Summary of O3 monitors meeting 40 CFR Part 58, Appendix A Precision
and Bias Goals.
Figure 3-1 7 Box plots of precision data by year (2005-2009) for all O3 monitors
reporting single-point QC check data to AQS.
Figure 3-1 8 Box plots of percent-difference data by year (2005-2009) for all
O3 monitors reporting single-point QC check data to AQS.
Figure 3-1 9 Box plots of RPD data by year for the co-located O3 monitors at two sites
in Missouri from 2006-2009.
Figure 3-20 Box plots of RPD data by year for all of the United States. Ozone sites
reporting single-point QC check data to AQS from 2005-2009.
Performance Specifications
Table 3-4 Performance specifications for O3 based in 40 CFR Part 53.
Monitor Calibration
Other Monitoring Technigues
Ambient O3 Network Design
Figure 3-21 U.S. O3 sites reporting data to AQS in 201 0.
Figure 3-22 U.S. Rural NCore, CASTNETand NPS POMS O3 sites in 2010.
nt Concentrations
Measurement Units, Metrics, and Averaging Times
3-63
3-63
3-66
3-66
3-67
3-67
3-68
3-69
3-69
3-70
3-70
3-71
3-75
3-77
3-79
3-80
3-80
3.6.2
Figure 3-23 Distribution in nation-wide year-round site-level correlations between daily
O3 metrics including 24-h avg, 1-h daily max and 8-h daily max using
AQS data, 2007-2009.
Spatial Variability
Figure 3-24
Table 3-5
Figure 3-25
Figure 3-26
Table 3-6
Table 3-7
Table 3-8
Figure 3-27
Figure 3-28
Table 3-9
Table 3-10
Figure 3-29
Figure 3-30
Figure 3-31
Figure 3-32
Figure 3-33
Reguired O3 monitoring time periods (ozone season) identified by
monitoring site.
Summary of O3 data sets originating from AQS.
Location of the 457 O3 monitors meeting the year-round data set
completeness criterion for all 3 years between 2007 and 2009.
Location of the 1,064 O3 monitors meeting the warm-season data set
completeness criteria for all 3 years between 2007 and 2009.
Nationwide distributions of O3 concentrations (ppb) from the year-round
data set.
Nationwide distributions of O3 concentrations (ppb) from the
warm-season data set.
Seasonally stratified distributions of 8-h daily max O3 concentrations
(ppb) from the year-round data set (2007-2009).
Highest monitor (by county) 3-year avg (2007-2009) of the 8-h daily max
O3 concentration based on the year-round data set (the top map) with
seasonal stratification (the four bottom maps).
Highest monitor (by county) 3-year avg (2007-2009) of the 8-h daily max
O3 concentration based on the warm-season data set (the top map) with
annual stratification (the three bottom maps).
Focus cities used in this and previous assessments.
City-specific distributions of 8-h daily max O3 concentrations (ppb) from
the warm-season data set (2007-2009).
Map of the Atlanta, Georgia, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways.
Map of the Boston, Massachusetts, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways.
Map of the Los Angeles, California, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways.
_3-82
_3-82
_3-83
_3-84
_3-85
_3-85
_3-87
_3-88
3-90
3-91
_3-92
_3-95
_3-96
_3-97
_3-97
3-98
Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Atlanta CSA. 3-100
Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Boston CSA. 3-100
-------
Figure 3-34
Figure 3-35
Figure 3-36
Figure 3-37
Figure 3-38
Figure 3-39
Figure 3-40
Figure 3-41
Figure 3-42
Table 3-11
Figure 3-43
Figure 3-44
Figure 3-45
Figure 3-46
Figure 3-47
3.6.3 Temporal Variability _
Figure 3-48
Figure 3-49
Figure 3-50
Figure 3-51
Figure 3-52
Figure 3-53
Figure 3-54
Figure 3-55
Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Los Angeles CSA. 3-101
Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the Atlanta CSA. 3-103
Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the Boston CSA.
Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the Los Angeles CSA.
Pair-wise monitor coefficient of divergence (COD) expressed as a
histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for 8-h daily max O3 in the Atlanta CSA. _
Pair-wise monitor coefficient of divergence (COD) expressed as a
histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for 8-h daily max O3 in the Boston CSA. _
Pair-wise monitor coefficient of divergence (COD) expressed as a
histogram (top), contour matrix (middle) and scatter plot versus distance
between monitors (bottom) for 8-h daily max O3 in the Los Angeles CSA._
Terrain map showing the location of two nearby AQS O3 monitoring sites
(red dots) along the western edge of the Los Angeles CSA. _
Terrain map showing the location of four AQS O3 monitoring sites (red
dots) located in or near the city limits in the center of the Boston CSA.
Rural focus areas.
Terrain map showing the location of five AQS O3 monitoring sites
(green/black stars) in Great Smoky Mountain National Park, NC-TN
(SMNP). _
Pair-wise monitor correlations (left) and coefficients of divergence (CODs)
(right) expressed as a histogram (top), contour matrix (middle) and
scatter plot vs. distance between monitors (bottom) for 8-h daily max O3
in Great Smoky Mountain National Park, NC-TN (SMNP). _
Terrain map showing the location of the AQS O3 monitoring site in Rocky
Mountain National Park, Colorado (black/green star) and the Denver,
Colorado, CSA (red dots) along with O3 monitoring sites used in the
Brodin et al. (2010) study (blue circles).
Terrain map showing the location of two AQS O3 monitoring sites
(black/green stars) in Sequoia National Park, CA.
3.6.4
Associations with Copollutants
Figure 3-56 Distribution of Pearson correlation coefficients for comparison of
8-h daily max O3 from the year-round data set with co-located 24-h avg
CO, SO2, NO2, PM10 and PM2.5 from AQS, 2007-2009.
_3-1 04
_3-1 05
_3-1 06
_3-1 07
_3-1 08
_3-1 1 1
Rural focus area site information, statistics and box plots for
8-h daily max O3 from AQS monitors meeting the warm-season data set
inclusion criteria within the rural focus areas.
_3-1 20
National 8-h daily max O3 trend and distribution across 870 U.S.
O3 monitors, 1998-2010 (annual 4th-highest 8-h daily max O3
concentrations in ppm). 3-121
National 1-h daily max O3 trend and distribution across 875 U.S.
O3 monitors, 1998-2010 (annual 2nd-highest 1-h daily max
O3 concentrations in ppm). 3-122
Trend in mean 8-h daily max O3 by region, 1998-2010 (mean of the
annual 4th-highest 8-h daily max O3 concentrations in ppm). 3-123
Trend in mean 1-h daily max O3 by region, 1998-2010 (mean of the
annual 2nd-highest 1-h daily max O3 concentrations in ppm). 3-124
Individual monitor 8-h daily max O3 design values displayed: (A) for the
2008-2010 period, and (B) as the change since the 2001-2003 period. 3-125
Individual monitor 1 -h daily max O3 design values displayed: (A) for the
2008-2010 period, and (B) as the change since the 2001-2003 period. 3-126
Diel patterns in 1-h avg O3 for Atlanta, Boston and Los Angeles between
2007 and 2009. 3-130
Diel patterns in 1 -h avg O3 for six rural focus areas between 2007 and
2009.
_3-1 33
_3-1 34
3-135
-------
Figure 3-57 Distribution of Pearson correlation coefficients for comparison of
8-h daily max O3 from the warm-season (May-Sept) data set with
co-located 24-h avg CO, SO2, NO2, PM10 and PM25 from AQS,
2007-2009. 3-136
3.7
Chapter Summary
3.7.1 Physical and Chemical Processes
3.7.2 Atmospheric Modeling
3.7.3 Background Concentrations
3.7.4 Monitoring
3.7.5 Ambient Concentrations
3-137
3-137
3-138
3-139
3-141
3-142
3.8 Supplemental Information on O3 Model Predictions 3-144
Figure 3-58 Comparison of time series of measurements of MDA8 O3 concentrations
at four CASTNET sites in the Northeast with GEOS-Chem predictions for
the base case and for the North American background case during
March-August, 2006. 3-145
Figure 3-59 Comparison of time series of measurements of MDA8 O3 concentrations
at four CASTNET sites in the Southeast with GEOS-Chem predictions for
the base case and for the North American background case during
March-August, 2006. 3-146
Figure 3-60 Comparison of time series of measurements of MDA8 O3 concentrations
at four CASTNET sites in the Upper Midwest with GEOS-Chem
predictions for the base case and for the North American background
case during March-August, 2006. 3-146
Figure 3-61 Comparison of time series of measurements of MDA8 O3 concentrations
at four CASTNET sites in the Intermountain West with GEOS-Chem
predictions for the base case and the North American background case
during March-August, 2006. 3-147
Figure 3-62 Comparison of time series of measurements of MDA8 O3 concentrations
at four CASTNET sites in the Intermountain West with GEOS-Chem
predictions for the base case and the North American background case
during March-August, 2006. 3-147
Figure 3-63 Comparison of time series of measurements of MDA8 O3 concentrations
at four CASTNET sites in the West with GEOS-Chem predictions for the
base case and the North American background case during March-
August, 2006. 3-148
Figure 3-64 Comparison of time series of measurements of MDA8 O3 concentrations
at three CASTNET sites and the Trinidad Head site in California with
GEOS-Chem predictions for the base case and the North American
background case during March-August, 2006. 3-148
Figure 3-65 Comparison of MDA8 O3 predicted using GEOS-Chem at 0.5° x 0.667°
(and 2° x 2.5° resolution; left figure only) with measurements at Mount
Bachelor, Oregon (left); and at Trinidad Head, California (right) from
March to August 2006. 3-149
Figure 3-66 Comparison of monthly mean (± 1 standard deviation) O3 calculated
GEOS-Chem (in red) with ozonesondes (in black) at Trinidad Head, CA
(top) and Boulder, Colorado (bottom) during April and August 2006. 3-149
Figure 3-67 A deep stratospheric O3 intrusion over California on May 28 to
May 29, 2010. 3-150
Figure 3-68 A deep stratospheric O3 intrusion over California on June 7 to
June 12, 2010. 3-151
Figure 3-69 Box plots showing maximum, interquartile range and minimum
O3 concentrations measured at CASTNET sites (black) in the Northeast
and predictions from GEOS-Chem at -50 x 50 km resolution (green) and
CAMx at 12 x 12 km resolution (blue) for May-August 2006. 3-152
Figure 3-70 Box plots showing maximum, interquartile range and minimum
O3 concentrations measured at CASTNET sites (black) in the Southeast
and predictions from GEOS-Chem at ~50 x 50 km resolution (green) and
CAMx at 12 x 12 km resolution (blue) for May-August 2006. 3-153
Figure 3-71 Box plots showing maximum, interquartile range and minimum
O3 concentrations measured at CASTNET sites (black) in the Central
U.S. and predictions from GEOS-Chem at ~50 x 50 km resolution (green)
and CAMx at 12 x 12 km resolution (blue) for May-August 2006. 3-154
Vlll
-------
Figure 3-72 Box plots showing maximum, interquartile range and minimum
O3 concentrations measured at CASTNET sites (black) in the Northern
Rockies and predictions from GEOS-Chem at -50 x 50 km resolution
(green) and CAMx at 12 x 12 km resolution (blue) for May-August 2006. 3-155
Figure 3-73 Box plots showing maximum, interquartile range and minimum
O3 concentrations measured at CASTNET sites (black) in the Southern
Rockies and predictions from GEOS-Chem at -50 x 50 km resolution
(green) and CAMx at 12 x 12 km resolution (blue) for May-August 2006. 3-156
Figure 3-74 Box plots showing maximum, interquartile range and minimum
O3 concentrations measured at CASTNET sites (black) in the West and
predictions from GEOS-Chem at -50 x 50 km resolution (green) and
CAMx at 12 x 12 km resolution (blue) for May-August 2006. 3-157
Figure 3-75 MDA8 O3 in surface air at Gothic, Colorado for March through August
2006. 3-158
3.9 Supplemental Figures of Observed Ambient O3 Concentrations 3-158
3.9.1 Ozone Monitor Maps for the Urban Focus Cities 3-158
Figure 3-76 Map of the Atlanta, Georgia, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-159
Figure 3-77 Map of the Baltimore, Maryland, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-159
Figure 3-78 Map of the Birmingham, Alabama, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-160
Figure 3-79 Map of the Boston, Massachusetts, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-160
Figure 3-80 Map of the Chicago, Illinois, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-161
Figure 3-81 Map of the Dallas, Texas, CSA including O3 monitor locations, population
gravity centers, urban areas, and major roadways. 3-161
Figure 3-82 Map of the Denver, Colorado, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-162
Figure 3-83 Map of the Detroit, Michigan, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-162
Figure 3-84 Map of the Houston, Texas, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-163
Figure 3-85 Map of the Los Angeles, California, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-163
Figure 3-86 Map of the Minneapolis, Minnesota, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-164
Figure 3-87 Map of the New York City, New York, CSA including O3 monitor
locations, population gravity centers, urban areas, and major roadways. 3-164
Figure 3-88 Map of the Philadelphia, Pennsylvania, CSA including O3 monitor
locations, population gravity centers, urban areas, and major roadways. 3-165
Figure 3-89 Map of the Phoenix, Arizona, CBSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-165
Figure 3-90 Map of the Pittsburgh, Pennsylvania, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-166
Figure 3-91 Map of the Salt Lake City, Utah, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-166
Figure 3-92 Map of the San Antonio, Texas, CBSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-167
Figure 3-93 Map of the San Francisco, California, CSA including O3 monitor
locations, population gravity centers, urban areas, and major roadways. 3-167
Figure 3-94 Map of the Seattle, Washington, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-168
Figure 3-95 Map of the St. Louis, Missouri, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways. 3-168
3.9.2 Ozone Concentration Box Plots for the Urban Focus Cities 3-169
Figure 3-96 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Atlanta, Georgia, CSA. 3-169
Figure 3-97 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Baltimore, Maryland, CSA. 3-170
-------
Figure 3-98 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Birmingham, Alabama, CSA. 3-170
Figure 3-99 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Boston, Massachusetts, CSA. 3-171
Figure 3-100 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Chicago, Illinois, CSA. 3-171
Figure 3-101 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Dallas, Texas, CSA. 3-172
Figure 3-102 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Denver, Colorado, CSA. 3-172
Figure 3-103 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Detroit, Michigan, CSA. 3-173
Figure 3-104 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Houston, Texas, CSA. 3-173
Figure 3-105 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Los Angeles, California, CSA. 3-174
Figure 3-106 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Minneapolis, Minnesota, CSA. 3-175
Figure 3-107 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
New York City, New York, CSA. 3-175
Figure 3-108 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Philadelphia, Pennsylvania, CSA. 3-176
Figure 3-109 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Phoenix, Arizona, CBSA. 3-176
Figure 3-110 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Pittsburgh, Pennsylvania, CSA. 3-177
Figure 3-111 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Salt Lake City, Utah, CSA. 3-177
Figure 3-112 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
San Antonio, Texas, CBSA. 3-178
Figure 3-113 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
San Francisco, California, CSA. 3-178
Figure 3-114 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
Seattle, Washington, CSA. 3-179
Figure 3-115 Site information, statistics and box plots for 8-h daily max O3 from AQS
monitors meeting the warm-season data set inclusion criteria within the
St. Louis, Missouri, CSA. 3-179
3.9.3 Ozone Concentration Relationships for the Urban Focus Cities 3-180
Figure 3-116 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the Atlanta, Georgia, CSA. 3-180
Figure 3-117 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the Baltimore, Maryland, CSA. 3-181
Figure 3-118 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the Birmingham, Alabama, CSA. 3-182
-------
Figure 3-119 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for8-h daily max O3 in the Boston, Massachusetts, CSA. 3-183
Figure 3-120 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for8-h daily max O3 in the Chicago, Illinois, CSA. 3-184
Figure 3-121 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for8-h daily max O3 in the Dallas, Texas, CSA. 3-185
Figure 3-122 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for8-h daily max O3 in the Denver, Colorado, CSA. 3-186
Figure 3-123 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for8-h daily max O3 in the Detroit, Michigan, CSA. 3-187
Figure 3-124 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for8-h daily max O3 in the Houston, Texas, CSA. 3-188
Figure 3-125 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the Los Angeles, California, CSA. 3-189
Figure 3-126 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the Minneapolis, Minnesota, CSA. 3-190
Figure 3-127 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the New York City, New York, CSA. 3-191
Figure 3-128 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the Philadelphia, Pennsylvania, CSA. 3-192
Figure 3-129 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the Phoenix, Arizona, CBSA. 3-193
Figure 3-130 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the Pittsburgh, Pennsylvania, CSA. 3-194
Figure 3-131 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the Salt Lake City, Utah, CSA. 3-195
Figure 3-132 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the San Antonio, Texas, CBSA. 3-196
Figure 3-133 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the San Francisco, California, CSA. 3-197
Figure 3-134 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the Seattle, Washington, CSA. 3-198
Figure 3-135 Pair-wise monitor correlations expressed as a histogram (top), contour
matrix (middle) and scatter plot versus distance between monitors
(bottom) for 8-h daily max O3 in the St. Louis, Missouri, CSA. 3-199
Figure 3-136 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Atlanta, Georgia, CSA. 3-200
Figure 3-137 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Baltimore, Maryland, CSA. 3-201
Figure 3-138 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Birmingham, Alabama,
CSA. 3-202
Figure 3-139 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Boston, Massachusetts,
CSA. 3-203
XI
-------
Figure 3-140 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for8-h daily max O3 in the Chicago, Illinois, CSA. 3-204
Figure 3-141 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Dallas, Texas, CSA. 3-205
Figure 3-142 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Denver, Colorado, CSA. 3-206
Figure 3-143 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Detroit, Michigan, CSA. 3-207
Figure 3-144 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Houston, Texas, CSA. 3-208
Figure 3-145 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Los Angeles, California,
CSA. 3-209
Figure 3-146 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Minneapolis, Minnesota,
CSA. 3-210
Figure 3-147 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the New York City, New York,
CSA. 3-211
Figure 3-148 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Philadelphia, Pennsylvania,
CSA. 3-212
Figure 3-149 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Phoenix. Arizona, CBSA. 3-213
Figure 3-150 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Pittsburgh, Pennsylvania,
CSA. 3-214
Figure 3-151 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Salt Lake City, Utah, CSA. 3-215
Figure 3-152 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the San Antonio, Texas, CBSA. 3-216
Figure 3-153 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the San Francisco, California,
CSA. 3-217
Figure 3-154 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Seattle, Washington, CSA. 3-218
Figure 3-155 Pair-wise monitor coefficient of divergence expressed as a histogram
(top), contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the St. Louis, Missouri, CSA. 3-219
3.9.4 Hourly Variations in O3 for the Urban Focus Cities 3-220
Figure 3-156 Diel patterns in 1 -h avg O3 for select CSAs between 2007 and 2009
using the year-round data set for the cold month/warm month comparison
(left half) and the warm-season data set for the weekday/weekend
comparison (right half). 3-221
Figure 3-157 Diel patterns in 1 -h avg O3 for select CSAs between 2007 and 2009
using the year-round data set for the cold month/warm month comparison
(left half) and the warm-season data set for the weekday/weekend
comparison (right half). 3-222
Xll
-------
Figure 3-158 Diel patterns in 1 -h avg O3 for select CSAs between 2007 and 2009
using the year-round data set for the cold month/warm month comparison
(left half) and the warm-season data set for the weekday/weekend
comparison (right half). 3-223
Figure 3-159 Diel patterns in 1 -h avg O3 for select CSAs/CBSAs between 2007 and
2009 using the year-round data set for the cold month/warm month
comparison (left half) and the warm-season data set for the
weekday/weekend comparison (right half). 3-224
Figure 3-160 Diel patterns in 1 -h avg O3 for select CSAs/CBSAs between 2007 and
2009 using the year-round data set for the cold month/warm month
comparison (left half) and the warm-season data set for the
weekday/weekend comparison (right half). 3-225
References 3-226
4 EXPOSURE TO AMBIENT OZONE 4-1
4.1 Introduction 4-1
4.2 General Exposure Concepts 4-1
4.3 Exposure Measurement 4-4
4.3.1 Personal Monitoring Techniques 4-4
4.3.2 Indoor-Outdoor Concentration Relationships 4-5
Table 4-1 Relationships between indoor and outdoor O3 concentration. 4-6
4.3.3 Personal-Ambient Concentration Relationships 4-12
Figure 4-1 Variation in hourly personal and ambient concentrations of O3 and PM2.s
in various microenvironments during daytime hours. 4-13
Table 4-2 Correlations between personal and ambient O3 concentration. 4-16
Table 4-3 Ratios of personal to ambient O3 concentration. 4-21
4.3.4 Co-exposure to Other Pollutants and Environmental Stressors 4-26
Figure 4-2 Correlations between 1-week concentrations of O3 and copollutants
measured near roadways. 4-29
4.4 Exposure-Related Metrics 4-30
4.4.1 Activity Patterns 4-30
Table 4-4 Mean fraction of time spent in outdoor locations by various age groups in
the NHAPS study. 4-31
Table 4-5 Mean ventilation rates (L/min) at different activity levels for different age
groups. 4-32
Figure 4-3 Distribution of time that NHAPS respondents spent in ten
microenvironments based on smoothed 1-min diary data. 4-33
4.4.2 Ozone-Averting Behavior 4-34
4.4.3 Population Proximity to Fixed-Site Ozone Monitors 4-36
Figure 4-4 Map of the Atlanta CSA including O3 monitor locations and major
roadways with respect to census block group population density
estimates for 2009. 4-37
Figure 4-5 Map of the Boston CSA including O3 monitor locations and major
roadways with respect to census block group population density
estimates for 2009. 4-38
Figure 4-6 Map of the Los Angeles CSA including O3 monitor locations and major
roadways with respect to census block group population density
estimates for 2009. 4-39
Table 4-6 Fraction of the 2009 population living within a specified distance of an O3
monitor in selected U.S. cities. 4-42
4.5 Exposure Modeling 4-43
Table 4-7 Characteristics of exposure modeling approaches. 4-44
4.5.1 Concentration Surface Modeling 4-44
4.5.2 Residential Air Exchange Rate Modeling 4-47
4.5.3 Microenvironment-Based Models 4-48
4.6 Implications for Epidemiologic Studies 4-50
4.6.1 Non-Ambient Ozone Exposure 4-51
4.6.2 Spatial and Temporal Variability 4-51
4.6.3 Exposure Duration 4-55
4.6.4 Relationship between Personal Exposure and Ambient Concentration 4-57
Xlll
-------
4.7
4.6.5 Exposure to Copollutants and Ozone Reaction Products
4.6.6 Averting Behavior
Figure 4-7 Adjusted asthma hospital admissions by age on lagged O3 by alert
status, ages 5-1 9 years old.
Figure 4-8 Adjusted asthma hospital admissions by age on lagged O3 by alert
status, ages 20-64 years old.
4.6.7 Exposure Estimation Methods in Epidemiologic Studies
Summary and Conclusions
References
DOSIMETRY, MODE OF ACTION, AND SPECIES HOMOLOGY
5.1
5.2
5.3
5.4
5.5
5.6
Introduction
Figure 5-1 Schematic of the O3 exposure and response pathway.
Human and Animal Ozone Dosimetry
5.2.1 Introduction
Figure 5-2 Representation of respiratory tract regions in humans.
Figure 5-3 Structure of lower airways with progression from the large airways to the
alveolar region.
5.2.2 Ozone Uptake
Table 5-1 Human respiratory tract uptake efficiency data.
Figure 5-4 Total O3 uptake efficiency as a function of breathing frequency at a
constant minute ventilation of 30 L/min.
Table 5-2 General adult human inhalation rates by activity levels.
Figure 5-5 Modeled effect of exercise on tissue dose of the LRT.
5.2.3 Ozone Reactions and Reaction Products
Figure 5-6 Schematic overview of O3 interaction with PUFA in ELF and lung cells.
Figure 5-7 Details of the O3 interaction with the airway ELF to form secondary
oxidation products.
Possible Pathways/Modes of Action
5.3.1 Introduction
5.3.2 Activation of Neural Reflexes
5.3.3 Initiation of inflammation
5.3.4 Alteration of Epithelial Barrier Function
5.3.5 Sensitization of Bronchial Smooth Muscle
5.3.6 Modification of Innate/Adaptive Immune System Responses
5.3.7 Airways Remodeling
5.3.8 Systemic Inflammation and Oxidative/Nitrosative Stress
5.3.9 Impaired Alveolar-Arterial Oxygen Transfer
5.3.10 Summary
Figure 5-8 The modes of action/possible pathways underlying the health effects
resulting from inhalation exposure to O3.
Interindividual Variability in Response
5.4.1 Dosimetric Considerations
5.4.2 Mechanistic Considerations
Figure 5-9 Some factors, illustrated in yellow, that likely contribute to the
interindividual variability in responses resulting from inhalation of O3.
Species Homology and Interspecies Sensitivity
5.5.1 Interspecies Dosimetry
Figure 5-1 0 Humans and animals are similar in the regional pattern of O3 tissue dose
distribution.
Figure 5-1 1 Oxygen-18 incorporation into different fractions of BALF from humans
and rats exposed to 0.4 and 2.0 ppm 18O3.
5.5.2 Interspecies Homology of Response
5.5.3 Summary
Chapter Summary
References
4-58
4-58
4-60
4-60
4-61
4-62
4-65
5-1
5-1
5-2
5-2
5-2
5-3
5-5
5-6
5-7
5-13
5-16
5-18
5-19
5-20
5-28
5-29
5-29
5-29
5-33
5-38
5-40
5-43
5-46
5-48
5-50
5-50
5-51
5-54
5-54
5-55
5-70
5-71
5-71
5-74
5-75
5-76
5-79
5-79
5-81
XIV
-------
6 INTEGRATED HEALTH EFFECTS OF SHORT-TERM OZONE EXPOSURE
6-1
6.1 Introduction
6.2 Respiratory Effects
6.2.1 Lung Function_
Table 6-1
Figure 6-1
Figure 6-2
Figure 6-3
Table 6-2
Figure 6-4
Table 6-3
Figure 6-5
Table 6-4
Figure 6-6
Table 6-5
Table 6-6
Table 6-7
Figure 6-7
Table 6-8
Figure 6-8
Table 6-9
Table 6-10
Table 6-11
Figure 6-9
Table 6-12
Table 6-13
Figure 6-10
Activity levels used in controlled exposures of healthy young adults to O3.
Cross-study comparison of mean O3-induced FEN/! decrements following
6.6 hours of exposure to O3.
Mean and upper percentile concentrations of O3 in epidemiologic studies
of lung function in adults with respiratory disease.
Mean and upper percentile concentrations of O3 in epidemiologic studies
of lung function in populations not restricted to individuals with asthma.
Percent change in FEN/! or FVC in association with ambient O3
concentrations in studies of children in the general population.
Percent change in FEVi or FVC in association with ambient O3
concentrations in studies of children in the general population presented
in Figure 6-9 plus others.
Associations between ambient O3 concentration and lung function in
studies of adults.
Comparison of O3-associated changes in lung function in single- and
copollutant models.
Table 6-14 Comparison of O3-associated changes in lung function in single- and
copollutant models for studies presented in Figure 6-10 plus others.
6.2.2 Airway Hyperresponsiveness
6.2.3 Pulmonary Inflammation, Injury and Oxidative Stress
Table 6-15 Mean and upper percentile O3 concentrations in epidemiologic studies of
biological markers of pulmonary inflammation and oxidative stress.
6-1
6-1
_6-3
6-7
6-8
Frequency distributions of FEVi decrements observed by Schelegle et al.
(2009) in young healthy adults (16 F, 15 M) following 6.6-hour exposures
to O3 or filtered air. 6-17
Proportion of individuals predicted to have greater than 10%, 15%, and
20% O3-induced FEN/! decrements a following 6.6-hour exposure to O3
with moderate exercise. 6-19
Mean and upper percentile O3 concentrations in epidemiologic studies of
lung function in populations with increased outdoor exposures. 6-30
Changes in FEN/! (ml) or PEF (mL/sec) in association with ambient O3
concentrations among children attending summer camp. 6-33
Changes in FEVi or PEF in association with ambient O3 concentrations
among children attending summer camp for studies presented in
Figure 6-4. 6-34
Percent change in FEV, in association with ambient O3 concentrations
among adults exercising outdoors. 6-36
Percent change in FEN/! in association with ambient O3 concentrations
among adults exercising outdoors for studies presented in Figure 6-5,
and among children exercising outdoors. 6-37
Percent change in lung function in association with ambient O3
concentrations among outdoor workers. 6-40
Percent change in FEN/! or FEV^FVC in association with ambient O3
concentrations among outdoor workers for studies presented in
Figure 6-6. 6-41
Associations between ambient O3 concentration and FEVi decrements in
different ranges of ambient O3 concentrations. 6-42
Mean and upper percentile concentrations of O3 in epidemiologic studies
of lung function in children with asthma. 6-43
Percent change in FEVi in association with ambient O3 concentrations
among children with asthma. 6-45
Percent change in FEN/! in association with ambient O3 concentrations
among children with asthma for studies presented in Figure 6-7 plus
others. 6-46
Percent change in PEF or FEF25-75% in association with ambient O3
concentrations among children with asthma. 6-48
Percent change in PEF or FEF25-75% in association with ambient O3
concentrations among children with asthma for studies presented in
Figure 6-8 plus others. 6-49
6-56
6-59
6-60
_6-61
_6-63
_6-67
_6-68
_6-72
_6-76
6-84
-------
Figure 6-11
Table 6-16
Table 6-17
Table 6-18
Percent change in exhaled nitric oxide (eNO) in association with ambient
O3 concentrations in populations with and without asthma.
Percent change in exhaled nitric oxide (eNO) in association with ambient
O3 concentrations in populations with and without asthma for studies
presented in Figure 6-11.
Associations between short-term ambient O3 exposure and biological
markers of pulmonary inflammation and oxidative stress.
Morphometric observations in non-human primates after acute O3
exposure.
6.2.4
Respiratory Symptoms and Medication Use
Table 6-19 Mean and upper percentile O3 concentrations in epidemiologic studies of
respiratory symptoms, medication use, and activity levels in children with
asthma.
Figure 6-12
Table 6-20
Figure 6-13
Table 6-21
Table 6-22
Table 6-23
Figure 6-14
Table 6-24
Table 6-25
6.2.5
6.2.6
6.2.7
Lung Host Defenses
Allergic and Asthma-Related Responses
Hospital Admissions, Emergency Department Visits, and Physicians Visits
Table 6-26 Mean and upper percentile concentrations of respiratory-related hospital
admission and emergency department (ED) visit studies evaluated.
Percent increase in respiratory hospital admissions from natural spline
models with 8 df/yr for a 40 ppb increase in 1-h max O3 concentrations
for each location of the APHENA study.
Figure 6-15
Table 6-27
Figure 6-16
Figure 6-17
Figure 6-18
Figure 6-19
Table 6-28
Figure 6-20
Corresponding effect estimates for Figure 6-15.
Estimated relative risks (RRs) of asthma hospital admissions for
8-h max O3 concentrations at lag 0-1 allowing for possible nonlinear
relationships using natural splines.
Risk ratio for respiratory ED visits and different O3 exposure metrics in
Atlanta, GA, from 1993-2004.
Loess C-R estimates and twice-standard error estimates from
generalized additive models for associations between 8-h max 3-day
average O3 concentrations and ED visits for pediatric asthma.
Percent increase in respiratory-related hospital admission and ED visits in
studies that presented all-year and/or seasonal results.
Corresponding Effect Estimates for Figure 6-19.
Table 6-29
6.2.8 Respiratory Mortality
6.2.9 Summary and Causal Determination
6.3 Cardiovascular Effects
Percent increase in respiratory-related hospital admissions and ED visits
for studies that presented single and copollutant model results.
Corresponding effect estimates for Figure 6-20.
6.3.1
6.3.2
Controlled Human Exposure
Epidemiology
6-85
6-86
6-87
_6-98
6-100
6-103
Associations between ambient O3 concentrations and respiratory
symptoms in children with asthma. 6-105
Associations between ambient O3 concentrations and respiratory
symptoms in children with asthma for studies presented in Figure 6-12. 6-106
Associations between ambient O3 concentrations and asthma medication
use. 6-110
Associations between ambient O3 concentrations and asthma medication
use for studies presented in Figure 6-13. 6-111
Mean and upper percentile O3 concentrations in epidemiologic studies of
respiratory symptoms and medication use in adults with respiratory
disease. 6-113
Mean and upper percentile O3 concentrations in epidemiologic studies of
respiratory symptoms in populations not restricted to individuals with
asthma. 6-115
Associations between ambient O3 concentrations and respiratory
symptoms in children in the general population. 6-116
Associations between ambient O3 concentrations and respiratory
symptoms in children in the general population for studies represented in
Figure 6-14. 6-117
Associations between ambient O3 concentrations and respiratory
symptoms in single- and copollutant models.
_6-120
_6-122
_6-128
_6-130
_6-132
_6-137
_6-138
_6-144
_6-146
_6-149
_6-153
_6-154
_6-156
_6-157
_6-158
_6-159
6-165
_6-165
6-168
XVI
-------
6.4
6.5
6.6
Table 6-30 Characterization of O3 concentrations (in ppb) from studies of
arrhythmias.
Table 6-31 Characterization of O3 concentrations (in ppb) from studies of heart rate
variability.
Figure 6-21 Odds ratio (95% confidence interval) for ischemic stroke by quintiles of
O3 exposure
Table 6-32 Characterization of O3 concentrations (in ppb) from studies of
biomarkers.
Table 6-33 Characterization of O3 concentrations (in ppb) from studies of blood
pressure.
Table 6-34 Characterization of O3 concentrations (in ppb) from studies of hospital
admissions and ED visits.
Figure 6-22 Effect estimate (95% Cl) per increment ppb increase in O3 for over all
cardiovascular ED visits or hospital admissions.
Table 6-35 Effect estimate (95% Cl) per increment ppb increase in O3 for overall
cardiovascular ED visits or hospital admissions in studies presented in
Figure 6-22.
Figure 6-23 Effect estimate (95% Cl) per increment ppb increase in O3 for congestive
heart failure ED visits or hospital admissions.
Table 6-36 Effect estimate (95% Cl) per increment ppb increase in O3 for congestive
heart failure ED visits or hospital admissions for studies in Figure 6-23.
Figure 6-24 Effect estimate (95% Cl) per increment ppb increase in O3 for ischemic
heart disease, coronary heart disease, myocardial infarction, and angina
pectoris ED visits or hospital admissions.
Table 6-37 Effect estimate (95% Cl) per increment ppb increase in O3 for ischemic
heart disease, coronary heart disease, myocardial infarction, and angina
pectoris ED visits or hospital admissions for studies presented in
Figure 6-24.
Figure 6-25 Effect estimate (95% Cl) per increment ppb increase in O3 for stroke ED
visits or hospital admissions.
Table 6-38 Effect estimate (95% Cl) per increment ppb increase in O3 for stroke ED
visits or hospital admissions for studies presented in Figure 6-25.
Figure 6-26 Effect estimate (95% Cl) per increment ppb increase in O3 for arrhythmia
and dysrhythmia ED visits or hospital admissions.
Table 6-39 Effect estimate (95% Cl) per increment ppb increase in O3 for arrhythmia
and dysrhythmia ED visits or hospital admissions for studies presented in
Figure 6-26.
6.3.3 Toxicology
Table 6-40 Characterization of study details for Section 6.3.3.
6.3.4 Summary and Causal Determination
Central Nervous System Effects
Table 6-41 Central nervous system and behavioral effects of short-term O3 exposure
in rats.
6.4.1 Neuroendocrine Effects
6.4.2 Summary and Causal Determination
Effects on Other Organ Systems
6.5.1 Effects on the Liver and Xenobiotic Metabolism
6.5.2 Effects on Cutaneous and Ocular Tissues
Mortality
6.6.1 Summary of Findings from 2006 O3 AQCD
6.6.2 Associations of Mortality and Short-Term O3 Exposure
Figure 6-27 Summary of mortality risk estimates for short-term O3 exposure and all-
cause (nonaccidental) mortality from all-year and summer season
analyses.
Table 6-42 Corresponding effect estimates for Figure 6-27.
Table 6-43 Range of mean and upper percentile O3 concentrations in previous and
recent multicity studies.
Table 6-44 Correlations between PM and O3 by season and region.
Figure 6-28 Scatter plots of O3 mortality risk estimates with versus without adjustment
for PMio in NMMAPS cities.
6-169
6-172
6-177
6-178
6-184
6-186
6-191
6-192
6-194
6-195
6-196
6-197
6-198
6-199
6-200
6-201
6-203
6-209
6-210
6-211
6-216
6-217
6-218
6-219
6-219
6-220
6-220
6-220
6-221
6-221
6-222
6-223
6-226
6-227
-------
Figure 6-29
Figure 6-30
Table 6-45
Table 6-46
Table 6-47
Figure 6-31
Table 6-48
Table 6-49
Figure 6-32
Figure 6-33
Table 6-50
Figure 6-34
Table 6-51
Figure 6-35
Table 6-52
Figure 6-36
Figure 6-37
Table 6-53
6.6.3 Summary and Causal
6.7 Overall Summary
Community-specific O3-mortality risk estimates for nonaccidental
mortality per 10 ppb increase in same-day 24-h average summertime O3
concentrations in single-pollutant models and copollutant models with
sulfate.
Percent increase in all-cause (nonaccidental) and cause-specific mortality
from natural spline models with 8 df/yr from the APHENA study for single-
and copollutant models.
Corresponding effect estimates for Figure 6-30.
Sensitivity of O3 risk estimates per 10 ug/m3 increase in 24-h average O3
concentrations at lag 0-1 to alternative methods for adjustment of
seasonal trend, for all-cause mortality using Berkey MLE and TLNISE
Hierarchical Models.
Additional percent change in O3-related mortality for individual-level
characteristics.
Ozone mortality risk estimates and community-specific characteristics,
U.S., 1987-2000.
Percent change in all-cause mortality, for all ages, associated with a
40ppb increase in 1 -h max O3 concentrations at Lag 0-1 at the 25th and
75th percentile of the center-specific distribution of selected effect
modifiers.
Percentage increase in daily mortality for a 10 ppb increase in
24-h average O3 concentrations during the previous week by geographic
region in the U.S., 1987-2000.
Community-specific Bayesian O3-mortality risk estimates in 98
U.S. communities.
Map of spatially dependent O3-mortality coefficients for 8-h max O3
concentrations using summer data.
Estimated effect of a 10 ppb increase in 8-h max O3 concentrations on
mortality during the summer months for single-day and distributed lag
models.
Estimated combined smooth distributed lag for 48 U.S. cities during the
summer months.
Estimated percent increase in cause-specific mortality (and 95% CIs) for
a 10-ug/m increase in 8-h daily max O3 during June-August.
Estimated combined smooth distributed lag in 21 European cities during
the summer (June-August) months.
Percent excess all-cause mortality per 10 ppb increase in daily
8-h max O3 on the same day, by season, month, and age groups.
Estimated combined C-R curve for nonaccidental mortality and
24-hour average O3 concentrations at lag 0-1 using the nonlinear (spline)
model.
Percent increase in cause-specific mortality.
Corresponding effect estimates for Figure 6-37.
Determination
_6-229
_6-231
_6-232
_6-234
_6-237
_6-239
_6-240
_6-242
_6-242
_6-243
_6-247
_6-248
_6-249
_6-251
6-252
Table 6-54 Summary of causal determinations for short-term exposures to O3._
References
_6-255
_6-259
_6-260
_6-261
6-264
_6-264
6-265
INTEGRATED HEALTH EFFECTS OF LONG-TERM OZONE EXPOSURE.
Introduction
7.1
7.2
Respiratory Effects_
7.2.1 Asthma
7.2.2
Figure 7-2 Ozone modifies the effect of TNF GG genotype on bronchitic symptoms
among children with asthma in the CHS.
Asthma Hospital Admissions and ED Visits
Figure 7-3 Ozone-asthma concentration-response relationship using the mean
concentration during the entire follow-up period for first asthma hospital
admission.
7.2.3 Pulmonary Structure and Function,
.7-1
. 7-1
. 7-2
7-3
Figure 7-1 Interaction of heme-oxygenase genetic variants and O3 level on the
Hazard Ratio (HR) of new-onset asthma in the 12 Children's Health Study
communities.
7-7
7-14
_7-16
7-17
XV111
-------
7.3
7.4
7.5
7.6
7.7
7.8
Table 7-1 Respiratory effects in nonhuman primates and rodents resulting from
long-term O3 exposure.
7.2.4 Pulmonary Inflammation, Iniury, and Oxidative Stress
7.2.5 Allergic Responses
7.2.6 Host Defense
7.2.7 Respiratory Mortality
7.2.8 Summary and Causal Determination
Table 7-2 Summary of selected key new studies examining annual O3 exposure
and respiratory health effects.
Table 7-3 Studies providing evidence concerning potential confounding by PM for
available endpoints.
Cardiovascular Effects
7.3.1 Cardiovascular Disease
Table 7-4 Characterization of study details for Section 7.3.1 .2.
7.3.2 Cardiovascular Mortality
7.3.3 Summary and Causal Determination
Reproductive and Developmental Effects
7.4.1 Effects on Sperm
7.4.2 Effects on Reproduction
7.4.3 Birth Weight
Figure 7-4 Birthweight deficit by decile of 24-h avg O3 concentration averaged over
the entire pregnancy compared with the decile group with the lowest O3
exposure.
Table 7-5 Brief summary of epidemiologic studies of birth weight.
7.4.4 Preterm Birth
Table 7-6 Brief summary of epidemiologic studies of preterm birth (PTB).
7.4.5 Fetal Growth
Table 7-7 Brief summary of epidemiologic studies of fetal growth.
7.4.6 Postnatal Growth
7.4.7 Birth Defects
Table 7-8 Brief summary of epidemiologic studies of birth defects.
7.4.8 Developmental Respiratory Effects
7.4.9 Developmental Central Nervous System Effects
7.4.10 Early Life Mortality
Table 7-9 Brief summary of infant mortality studies.
Table 7-1 0 Summary of key reproductive and developmental toxicological studies.
7.4.1 1 Summary and Causal Determination
Central Nervous System Effects
7.5.1 Effects on the Brain and Behavior
Table 7-1 1 Central nervous system effects of long-term O3 exposure in rats.
7.5.2 Summary and Causal Determination
Carcinogenic and Genotoxic Potential of Ozone
7.6.1 Introduction
7.6.2 Lung Cancer Incidence and Mortality
7.6.3 DNA Damage
7.6.4 Summary and Causal Determination
Mortality
Figure 7-5 Adjusted O3-mortality relative risk estimates (95% Cl) by time period of
analysis per subject-weighted mean O3 concentration in the Cancer
Prevention Study II by the American Cancer Society.
Table 7-1 2 Relative risk (and 95% Cl) of death attributable to a 1 0-ppb change in the
ambient O-, concentration.
7.7.1 Summary and Causal Determination
Overall Summary
Table 7-13 Summary of causal determinations for long-term exposures to O3.
References
7-25
7-27
7-29
7-30
7-31
7-31
7-33
7-35
7-36
7-36
7-39
7-39
7-40
7-40
7-42
7-44
7-45
7-46
7-48
7-49
7-53
7-54
7-57
7-57
7-58
7-61
7-61
7-65
7-67
7-71
7-73
7-74
7-75
7-75
7-79
7-79
7-80
7-80
7-82
7-82
7-85
7-85
7-86
7-89
7-89
7-90
7-91
7-92
XIX
-------
8 POPULATIONS POTENTIALLY AT INCREASED RISK FOR OZONE-RELATED
HEALTH EFFECTS
8.1
8.2
8.3
8.4
8.5
Table 8-1 Classification of Evidence for Potential At-Risk Factors.
Genetic Factors
Table 8-2 Summaries of results from epidemiologic and controlled human
exposures studies of modification by genetic variants.
Table 8-3 Summaries of results from animal toxicology studies of modification by
genetic variants.
Pre-existing Disease/Conditions
Table 8-4 Prevalence of respiratory diseases, cardiovascular diseases, and
diabetes among adults by age and region in the U.S.
8.2.1 Influenza/Infections
8.2.2 Asthma
Table 8-5 Prevalence of asthma by age in the U.S.
8.2.3 Chronic Obstructive Pulmonary Disease (COPD)
8.2.4 Cardiovascular Disease (CVD)
8.2.5 Diabetes
8.2.6 Hyperthyroidism
Sociodemographic Factors
8.3.1 Lifestage
8.3.2 Sex
8.3.3 Socioeconomic Status
8.3.4 Race/Ethnicity
Behavioral and Other Factors
8.4.1 Diet
8.4.2 Obesity
8.4.3 Smoking8-32
8.4.4 Outdoor Workers
8.4.5 Air Conditioning Use
Summary
Table 8-6 Summary of evidence for potential increased risk of O3-related health
effects.
References
ENVIRONMENTAL EFFECTS: OZONE EFFECTS ON VEGETATION
AND ECOSYSTEMS
9.1
9.2
9.3
Introduction
Figure 9-1 An illustrative diagram of the major endpoints that O3 may affect in plants
and ecosystems.
Experimental Exposure Methodologies
9.2.1 Introduction
9.2.2 "Indoor," Controlled Environment, and Greenhouse Chambers
9.2.3 Field Chambers
9.2.4 Plume and FACE-Type Systems
9.2.5 Ambient Gradients
9.2.6 Comparative Studies
Mechanisms Governing Vegetation Response to Ozone
9.3.1 Introduction
9.3.2 Ozone Uptake into the Leaf
Figure 9-2 Ozone uptake from the atmosphere (A), and The anatomy of a dicot leaf
(B).
Figure 9-3 Possible reactions of O3 within water.
Figure 9-4 The Crigee mechanism of O3 attack of a double bond.
9.3.3 Cellular to Systemic Responses
Figure 9-5 Composite diagram of major themes in the temporal evolution of the
genetic response to O3 stress.
8-1
8-3
8-3
8-5
8-7
8-10
8-11
8-11
8-12
8-12
8-15
8-15
8-17
8-17
8-18
8-18
8-24
8-26
8-28
8-30
8-30
8-31
8-33
8-34
8-35
8-36
8-38
9-1
9-1
9-3
9-3
9-3
9-4
9-4
9-6
9-7
9-8
9-10
9-10
9-11
9-14
9-15
9-15
9-16
9-21
XX
-------
Figure 9-6 The oxidative cell death cycle. 9-24
9.3.4 Detoxification 9-24
9.3.5 Effects on Primary and Secondary Metabolism 9-28
9.3.6 Summary 9-34
9.4 Nature of Effects on Vegetation and Ecosystems 9-36
9.4.1 Introduction 9-36
9.4.2 Visible Foliar Injury and Biomonitoring 9-38
9.4.3 Growth, Productivity and Carbon Storage in Natural Ecosystems 9-42
Table 9-1 Ozone effects on plant reproductive processes. 9-47
Table 9-2 Comparison of models used to simulate the ecological consequences of
O3 exposure. 9-50
Table 9-3 Modeled effects of O3 on primary production, C exchange,
and C sequestration. 9-56
9.4.4 Crop Yield and Quality in Agricultural Systems 9-57
Table 9-4 Summary of recent studies of O3 effects on crops (exclusive of growth
and yield). 9-64
Table 9-5 Modeled effects of O3 on crop yield loss at regional and global scales. 9-67
9.4.5 Water Cycling 9-67
Figure 9-7 The potential effects of O3 exposure on water cycling. 9-68
9.4.6 Below-Ground Processes 9-71
Figure 9-8 Conceptual diagram showing where O3 alters C, water and nutrient flow
in a tree-soil system, including transfer between biotic and abiotic
components below ground that influence soil physical and chemical
properties. 9-72
Table 9-6 The effect of elevated O3 on leaf/litter nutrient concentrations. 9-74
Table 9-7 The temporal variation of ecosystem responses to O3 exposure at Aspen
FACE site 9-77
9.4.7 Community Composition 9-80
9.4.8 Factors that Modify Functional and Growth Response 9-84
Table 9-8 Response of plants to the interactive effects of elevated O3 exposure and
nitrogen enrichment. 9-89
9.4.9 Insects and Other Wildlife 9-92
9.5 Effects-based Air Quality Exposure Indices and Dose Modeling 9-98
9.5.1 Introduction 9-98
9.5.2 Description of Exposure Indices Available in the Literature 9-99
Figure 9-9 Diagrammatic representation of several exposure indices illustrating how
they weight concentration and accumulate exposure. 9-100
9.5.3 Important Components of Exposure Indices 9-104
Figure 9-10 Trends in May to September: 12-hour SUM06, Peak 1 -hour O3
concentration and number of daily exceedances of 95 ppb for the
Crestline site in 1963 to 1999; in relation to trends in mean daily
maximum temperature for Crestline and daily reactive organic gases
(ROG) and oxides of nitrogen (NOX) for San Bernardino County. 9-107
Figure 9-11 The number of hourly average concentrations between 50 and 89 ppb for
the period 1980-2000 for the Crestline, San Bernardino County, CA,
monitoring site. 9-108
Figure 9-12 Diurnal (a) conductance through boundary layer and stomata (gbs), (b)
ozone concentration, and leaf-level stomatal O3 flux (FstOI) in control
plots from mid-June through August, in (c) 2004 and (d) 2005 in the
Aspen FACE experiment. 9-111
Figure 9-13 Maximum 3-month, 12-h W126 plotted against maximum 6-month,
12-h W126. 9-113
9.5.4 Ozone Uptake/Dose Modeling for Vegetation 9-114
9.5.5 Summary 9-116
9.6 Ozone Exposure-Plant Response Relationships 9-117
9.6.1 Introduction 9-117
9.6.2 Estimates of Crop Yield Loss and Tree Seedling Biomass Loss in the
1996 and 2006 Ozone AQCDs 9-120
Figure 9-14 Quantiles of predicted relative yield loss for 34 NCLAN crop experiments. 9-122
Figure 9-15 Quantiles of predicted relative yield loss for 4 crop species in NCLAN
experiments. 9-123
XXI
-------
Figure 9-16
Figure 9-17
Table 9-9
Table 9-10
Table 9-11
Quantiles of predicted relative biomass loss for 49 studies of 11 tree
species in NHEERL/WED experiments.
Quantiles of predicted relative biomass loss for 4 tree species in
NHEERL/WED experiments.
Ozone exposures at which 10 and 20% yield loss is predicted for 50 and
75% of crop species.
Ozone exposures at which 10 and 20% yield loss is predicted for 50 and
75% of crop species (Draughted versus Watered conditions).
Ozone exposures at which 10 and 20% biomass loss is predicted for 50
and 75% of tree species.
9.6.3
Validation of 1996 and 2006 Ozone AQCD Models and Methodology Using the 90-day 12-h
W126 and Current FACE Data
Table 9-12
Table 9-13
Figure 9-18
Figure 9-19
Table 9-14
Table 9-15
Figure 9-20
Comparison between above-ground biomass observed in Aspen FACE
experiment in 6 year and biomass predicted by the median composite
function based on NHEERL/WED.
9.7
9-124
9-125
9-126
9-126
9-127
9-127
Comparison between change in yield observed in the SoyFACE
experiment between elevated and ambient O3, and change predicted at
the same values of O3 by the median composite function for NCLAN. 9-130
Comparison between yield observed in the SoyFACE experiment and
yield predicted at the same values of O3 by the median composite
function for NCLAN. 9-130
Comparison of yield observed in SoyFACE experiment in a given year
with yield predicted by the median composite function based on NCLAN. 9-131
Comparison of composite functions for the quartiles of 7 curves for 7
genotypes of soybean grown in the SoyFACE experiment, and for the
quartiles of 11 curves for 5 genotypes of soybean grown in the NCLAN
project. 9-132
Comparison between change in above-ground biomass elevated and
ambient O3 in Aspen FACE experiment in 6 year, and change predicted
at the same values of O3 by the median composite function for
NHEERL/WED. 9-134
Comparison between above-ground biomass observed in Aspen FACE
experiment in 6 year and biomass predicted by the median composite
function based on NHEERL/WED. 9-134
9-135
9.6.4
Summ
;nces
Figure 9-21
Table 9-1 6
Summary
Table 9-1 7
Table 9-1 8
ary and Conclusions
Table 9-1 9
Above-ground biomass for one genotype of cottonwood grown in seven
locations for one season in 3 years.
Meta-analyses of growth or yield studies published since 2005.
Summary of studies of effects of O3 exposure on growth and yield of
agricultural crops.
Summary of studies of effects of O3 exposure on growth of natural
vegetation.
Summary of O3 causal determinations for vegetation and ecosystem
effects.
9-137
9-138
9-140
9-141
9-145
9-147
9-148
9-149
10 THE ROLE OF TROPOSPHERIC OZONE IN CLIMATE CHANGE AND UV-B
SHIELDING EFFECTS
10.1 Introduction
10.2 Physics of the Earth's Radiation Budget
Figure 10-1 Diagram of the factors that determine human exposure to ultraviolet
radiation.
10.3 Effects of Tropospheric O3 on Climate_
10.3.1 Background
10.3.2 Climate Change Evidence and the Influence of Tropospheric O3
Figure 10-2 Schematic illustrating the effects of tropospheric O3 on climate._
10.3.3 Factors that Influence the Effect of Tropospheric O3 on Climate_
.10-1
_ 10-1
10-1
_10-3
10-4
_10-4
_10-5
10-8
Figure 10-3 Global average radiative forcing (RF) estimates and uncertainty ranges in
2005 for anthropogenic CO2, CH4, O3, and other important agents and
mechanisms. 10-9
10-10
XX11
-------
10.3.4 Competing Effects of O3 Precursors on Climate 10-15
Figure 10-4 Components of radiative forcing for emissions of principal gases,
aerosols, aerosol precursors, and other changes. 10-17
10.3.5 Calculating Radiative Forcing and Climate Response to Past Trends in
Tropospheric O3 Concentrations 10-18
Figure 10-5 Ensemble average 1900-2000 radiative forcing and surface temperature
trends (°C per century) in response to tropospheric O3 concentration
changes. 10-19
10.3.6 Calculating Radiative Forcing and Climate Response to Future Trends in
Tropospheric O3 Concentrations 10-19
Table 10-1 Changes in anthropogenic emissions, CH4 and tropospheric O3
concentrations between 2000 and 2030, and the associated tropospheric
O3 radiative forcing for three scenarios. 10-22
Figure 10-6 Global mean radiative forcing estimates calculated by a set of models for
the 2000-2100 change in tropospheric O3 concentrations. 10-24
10.4 UV-B Shielding Effects and Tropospheric O3 10-25
10.4.1 Background 10-25
10.4.2 Human Exposure and Susceptibility to Ultraviolet Radiation 10-25
10.4.3 Human Health Effects due to UV-B Radiation 10-26
10.4.4 Ecosystem and Materials Damage Effects Due to UV-B Radiation 10-27
10.4.5 UV-B Shielding Effects Associated with Changes in Tropospheric O3 Concentrations 10-28
10.5 Summary and Causal Determinations 10-30
10.5.1 Summary of the Effects of Tropospheric O3 on Climate 10-30
10.5.2 Summary of UV-B Related Effects on Human Health, Ecosystems, and Materials
Relating to Changes in Tropospheric O3 Concentrations 10-31
10.5.3 Summary of O3 Causal Determinations 10-32
Table 10-2 Summary of O3 causal determinations for climate and UV-B
shielding effects. 10-32
References 10-33
XX111
-------
OZONE PROJECT TEAM
Executive Direction
Dr. John Vandenberg (Director)—National Center for Environmental Assessment-RTF
Division, Office of Research and Development, U.S. Environmental Protection
Agency, Research Triangle Park, NC
Ms. Debra Walsh (Deputy Director)—National Center for Environmental Assessment-
RTF Division, Office of Research and Development, U.S. Environmental Protection
Agency, Research Triangle Park, NC
Dr. Mary Ross (Branch Chief)—National Center for Environmental Assessment, Office
of Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Scientific Staff
Dr. James Brown (O3 Team Leader)—National Center for Environmental Assessment,
Office of Research and Development, U.S. Environmental Protection Agency,
Research Triangle Park, NC
Dr. Christal Bowman—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Barbara Buckley—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Ms. Ye Cao—Oak Ridge Institute for Science and Education, National Center for
Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Allen Davis—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. Jean-Jacques Dubois—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Steven J. Dutton—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Jeffrey Herrick—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. Erin Hines—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Doug Johns—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
XXIV
-------
Dr. Dennis Kotchmar—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Meredith Lassiter—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Lingli Liu— Oak Ridge Institute for Science and Education, National Center for
Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Thomas Long—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. Thomas Luben—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. Qingyu Meng— Oak Ridge Institute for Science and Education, National Center for
Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Kristopher Novak—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Elizabeth Oesterling Owens—National Center for Environmental Assessment, Office
of Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Molini Patel—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. Joseph P. Pinto—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Ms. Joann Rice—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Jason Sacks—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. Lisa Vinikoor-Imler—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
XXV
-------
Technical Support Staff
Mr. Kenneth J. Breito-Senior Environmental Employment Program, National Center for
Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Gerald Gurevich—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Mr. Ryan Jones—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Ellen Lorang—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Mr. J. Sawyer Lucy-Student Services Authority, National Center for Environmental
Assessment, Office of Research and Development, U.S. Environmental Protection
Agency, Research Triangle Park, NC
Ms. Deborah Wales—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Mr. Richard N. Wilson-National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Ms. Barbara Wright—Senior Environmental Employment Program, National Center for
Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
XXVI
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
Authors
Dr. James Brown (O3 Team Leader)—National Center for Environmental Assessment,
Office of Research and Development, U.S. Environmental Protection Agency,
Research Triangle Park, NC
Dr. Christal Bowman—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Barbara Buckley—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Ms. Ye Cao—Oak Ridge Institute for Science and Education, National Center for
Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Maggie Clark—Department of Environmental and Radiological Health Sciences,
Colorado State University, Fort Collins, CO
Dr. Jean-Jacques Dubois—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Steven J. Dutton—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Arlene M. Fiore—Department of Earth and Environmental Sciences, Columbia
University and Lamont-Doherty Earth Observatory, Palisades, NY
Dr. Kelly Gillespie— Donald Danforth Plant Science Center, St. Louis, MO
Dr. Terry Gordon—Department of Environmental Medicine, New York University
School of Medicine, Tuxedo, NY
Dr. Barren Henderson—Oak Ridge Institute for Science and Education, National
Exposure Research Lab, Office of Research and Development, U.S. Environmental
Protection Agency, Research Triangle Park, NC
Dr. Jeffrey Herrick—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. Erin Hines—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Kazuhiko Ito—Department of Environmental Medicine, New York University
School of Medicine, Tuxedo, NY
Dr. Doug Johns—National Center for Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
XXVH
-------
Dr. Dennis Kotchmar—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Meredith Lassiter—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Lingli Liu— Oak Ridge Institute for Science and Education, National Center for
Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Thomas Long—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. Thomas Luben—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. Loretta J. Mickley—School of Engineering & Applied Sciences, Harvard University,
Cambridge, MA
Dr. Kristopher Novak—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Elizabeth Oesterling Owens—National Center for Environmental Assessment, Office
of Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Molini Patel—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. Jennifer Peel—Department of Environmental and Radiological Health Sciences,
Colorado State University, Fort Collins, CO
Dr. Joseph Pinto—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. Edward Postlethwait—Department of Environmental Health Sciences, School of
Public Health, University of Alabama at Birmingham, Birmingham, AL
Ms. Joann Rice—on detail to the National Center for Environmental Assessment, Office
of Research and Development, from the Office of Air Quality Planning and Standards,
Office of Air and Radiation, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Mr. Jason Sacks—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. George Thurston—Department of Environmental Medicine, New York University
School of Medicine, Tuxedo, NY
XXVlll
-------
Dr. Lisa Vinikoor-Imler—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. Cosima Wiese—Department of Biology, Misericordia University, Dallas, PA
Contributors
Mr. Brian Adams—Oak Ridge Institute for Science and Education, National Center for
Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Halil Cakir—Oak Ridge Institute for Science and Education, National Center for
Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Allen Davis—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Mr. Mark Evangelista—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Jay Haney—ICF International, San Rafael, CA
Dr. E. Henry Lee—National Health and Environmental Effects Research Laboratory,
U.S. Environmental Protection Agency, Corvallis, OR
Dr. Meiyun Lin—Atmospheric and Oceanic Sciences Program, Princeton University,
Princeton, NJ
Dr. Qingyu Meng—Oak Ridge Institute for Science and Education, National Center for
Environmental Assessment, Office of Research and Development, U.S.
Environmental Protection Agency, Research Triangle Park, NC
Mr. David Mintz—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Tom Myers—ICF International, San Rafael, CA
Dr. Jennifer Nichols ~ Oak Ridge Institute for Science and Education, National Center
for Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Michelle Oakes—Oak Ridge Institute for Science and Education, National Center for
Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Mr. Jacob T. Oberman—Center for Sustainability and the Global Environment (SAGE),
Nelson Institute for Environmental Studies, University of Wisconsin-Madison,
Madison, WI
Mr. Mark Schmidt—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Kaylyn Siporin—Oak Ridge Institute for Science and Education, National Center for
Environmental Assessment, Office of Research and Development, U.S.
Environmental Protection Agency, Research Triangle Park, NC
XXIX
-------
Dr. Gail Tonne sen, Region 8, U.S. Environmental Protection Agency, Denver, CO
Dr. Huiquin Wang—School of Engineering and Applied Science, Harvard University,
Cambridge, MA
Mr. Benjamin Wells—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Adrien Wilkie—Oak Ridge Institute for Science and Education, National Center for
Environmental Assessment, Office of Research and Development, U.S.
Environmental Protection Agency, Research Triangle Park, NC
Ms. Brianna Young—Oak Ridge Institute for Science and Education, National Center for
Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Lin Zhang—School of Engineering and Applied Science, Harvard University,
Cambridge, MA
Reviewers
Dr. Christian Andersen—National Health and Environmental Effects Research
Laboratory, U.S. Environmental Protection Agency, Corvallis, OR
Ms. Lea Anderson—Office of General Counsel, U.S. Environmental Protection Agency,
Washington, D.C.
Dr. Susan Anenberg—Office of Air Quality Planning and Standards, U.S. Environmental
Protection Agency, Washington, D.C.
Dr. Robert Arnts—National Exposure Research Laboratory, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. John Balmes— Department of Medicine, University of California, San Francisco and
School of Public Health, University of California, Berkeley, CA
Dr. Lisa Baxter—National Exposure Research Laboratory, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Souad Benromdhane—Office of Air Quality Planning and Standards, Office of Air
and Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Fitzgerald Booker—USDA-ARS Plant Science Research Unit, Raleigh, NC
Dr. Michael Breen—National Exposure Research Laboratory, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Philip Bromberg—School of Medicine, University of North Carolina, Chapel Hill,
NC
Dr. Kent Burkey—USDA-ARS Plant Science Research Unit, Raleigh, NC
Dr. David DeMarini—National Health and Environmental Effects Research Laboratory,
Office of Research and Development, U.S. Environmental Protection Agency,
Research Triangle Park, NC
Dr. Russ Dickerson—Department of Atmospheric and Oceanic Science, University of
Maryland, College Park, MD
XXX
-------
Mr. Patrick Dolwick—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Aimen Farraj—National Health and Environmental Effects Research Laboratory,
Office of Research and Development, U.S. Environmental Protection Agency,
Research Triangle Park, NC
Dr. Arlene Fiore—NOAA/Geophysical Dynamics Laboratory, Princeton, NJ
Dr. Ian Gilmour—National Health and Environmental Effects Research Laboratory,
Office of Research and Development, U.S. Environmental Protection Agency,
Research Triangle Park, NC
Dr. Stephen Graham—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Tara Greaver—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. Gary Hatch—National Health and Environmental Effects Research Laboratory,
Office of Research and Development, U.S. Environmental Protection Agency,
Research Triangle Park, NC
Dr. Bryan Hubbel—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Kristin Isaacs—National Exposure Research Laboratory, Office of Research and
Development, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Scott Jenkins—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Karl Jensen—National Health and Environmental Effects Research Laboratory,
Office of Research and Development, U.S. Environmental Protection Agency,
Research Triangle Park, NC
Dr. Urmila Kodavanti—National Health and Environmental Effects Research Laboratory.
Office of Research and Development, U.S. Environmental Protection Agency,
Research Triangle Park, NC
Dr. Petros Koutrakis—Department of Environmental Health, Harvard School of Public
Health, Boston, MA
Mr. John Langstaff—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Christopher Lau—National Health and Environmental Effects Research Laboratory,
Office of Research and Development, U.S. Environmental Protection Agency,
Research Triangle Park, NC
Mr. Gary Lear— Office of Administration and Policy, Office of Air and Radiation, U.S.
Environmental Protection Agency, Washington, DC
Dr. Morton Lippmann—Nelson Institute of Environmental Medicine, New York
University, Tuxedo, NY
Dr. Karen Martin—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
XXXI
-------
Ms. Connie Meacham—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Mr. David Mintz—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Pradeep Raj an—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. John Rogers—National Health and Environmental Effects Research Laboratory,
Office of Research and Development, U.S. Environmental Protection Agency,
Research Triangle Park, NC
Ms. Vicki Sandiford—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Ms. Susan Stone—Office of Air Quality Planning and Standards, Office of Air and
Radiation, U.S. Environmental Protection Agency, Research Triangle Park, NC
Dr. Gail Tonne sen, Region 8, U.S. Environmental Protection Agency, Denver, CO
Dr. John Vandenberg—National Center for Environmental Assessment, Office of
Research and Development, U.S. Environmental Protection Agency, Research
Triangle Park, NC
Dr. James G. Wagner—Department of Pathobiology and Diagnostic Investigation,
Michigan State University, East Lansing, MI
Ms. Debra Walsh—National Center for Environmental Assessment, Office of Research
and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC
Dr. Jason West—Department of Environmental Sciences & Engineering, University of
North Carolina, Chapel Hill, NC
XXXH
-------
CLEAN AIR SCIENTIFIC ADVISORY COMMITTEE OZONE
NAAQS REVIEW PANEL
Chair of the Environmental Protection Agency's Clean Air Scientific Advisory Committee
Dr. H. Christopher Frey*, Department of Civil, Construction and Environmental
Engineering, College of Engineering, North Carolina State University, Raleigh, NC
Chair of the Ozone Review Panel
Dr. H. Christopher Frey*, Department of Civil, Construction and Environmental
Engineering, College of Engineering, North Carolina State University, Raleigh, NC
Members
Dr. George A. Allen*, Northeast States for Coordinated Air Use Management
(NESCAUM), Boston, MA
Professor Ed Avol, Department of Preventive Medicine, Keck School of Medicine,
University of Southern California, Los Angeles, CA
Dr. John Bailar, The National Academies, Washington, D.C.
Dr. Michelle Bell, School of Forestry & Environmental Studies, Yale University, New
Haven, CT
Dr. Joseph D. Brain*, Department of Environmental Health, Harvard School of Public
Health, Harvard University, Boston, MA
Dr. David Chock, Independent Consultant, Bloomfield Hills, MI
Dr. Ana Diez-Roux*, Department of Epidemiology, University of Michigan School of
Public Health; Ann Arbor, MI
Dr. William Michael Foster, Division of Pulmonary, Allergy, and Critical Care Medicine,
Duke University Medical Center, Durham, NC
Dr. Judith Graham, Independent Consultant, Pittsboro, NC
Dr. David A. Grantz, College of Natural and Agricultural Sciences, Air Pollution
Research Center, University of California Riverside, Parlier, CA
Dr. Jack Harkema, Center for Integrated Toxicology, Michigan State University, East
Lansing, MI
Dr. Daniel Jacob, Atmospheric Chemistry and Environmental Engineering, Harvard
University, Cambridge, MA
Dr. Steven Kleeberger, National Institute of Environmental Health Sciences, National
Institutes of Health, Research Triangle Park, NC
Dr. Frederick J. Miller, Independent Consultant, Gary, NC
Dr. Howard Neufeld, Department of Biology, Appalachian State University, Boone, NC
XXXlll
-------
Dr. Armistead (Ted) Russell*, Department of Civil and Environmental Engineering,
Georgia Institute of Technology, Atlanta, GA
Dr. Jonathan M. Samet**, Department of Preventive Medicine at the Keck School of
Medicine, and Director of the Institute for Global Health at the University of Southern
California, Los Angeles, CA
Dr. Helen Suh*, Environmental Health, National Opinion Research Corporation (NORC)
at the University of Chicago, West Newton, MA
Dr. James Ultman, Department of Chemical Engineering, Pennsylvania State University,
University Park, PA
Dr. Sverre Vedal, Department of Environmental and Occupational Health Sciences,
School of Public Health and Community Medicine, University of Washington, Seattle,
WA
Dr. Kathleen Weathers*, Gary Institute of Ecosystem Studies, Millbrook, NY
Dr. Peter Woodbury, Department of Crop and Soil Sciences, Cornell University, Ithaca,
NY
* Members of the statutory Clean Air Scientific Advisory Committee (CASAC)
appointed by the EPA Administrator
** Immediate Past CASAC Chair, and Immediate Past Ozone Review Panel Chair
Science Advisory Board Staff
Dr. Holly Stallworth, Designated Federal Officer, U.S. Environmental Protection
Agency, Mail Code 1400R, 1300 Pennsylvania Avenue, NW, Washington, DC,
20004, Phone: 202-564-2073, Email: stallworth.holly@epa.gov
XXXIV
-------
ACRONYMS AND ABBREVIATIONS
129 mouse strain (12931/SvlmJ)
a alpha, ambient exposure factor
a-ATD alpha 1-antitrypsin deficiency
a-SMA alpha-smooth muscle actin
a-tocopherol alpha-tocopherol
a-TOH alpha tocopherol
a air exchange rate of the
microenvironment
A2 climate scenario in IPCC
AADT annual average daily traffic
A1 B climate scenario in IPCC
ABA abscisic acid
ABI abscisic acid insensitive
ABI2 abscisic acid insensitive
Arabidopsis mutant
A1c glycosylated hemoglobin blood
test
Ach acetylcholine
ACM (Harvard University) Atmospheric
Chemistry Modeling (Group)
ACS American Cancer Society
ACS-CPSII ACS Cancer Prevention Study II
ADC arginine decarboxylase
ADSP Adirondack State Park, NY
AER air exchange rate
AH2 ascorbic acid; ascorbate
AHR airway(s) hyperresponsiveness,
airway(s) hyperreactivity
AhR aryl hydrocarbon receptor
AHSMOG (California Seventh Day) Adventist
Heath and Smog (Study)
Al alveolar interstitial
AIC(s) Akaike's information criterion
AIRS Aerometric Information Retrieval
System; Atmospheric Infrared
Sounder (instrument)
A/J mouse strain
Ala-9Val genotype associated with
Manganese superoxide dismutase
(MnSOD) gene
AM alveolar macrophage(s)
ANF atrial natriuretic factor
AOT20 seasonal sum of the difference
between an hourly concentration
at the threshold value of 20 ppb,
minus the threshold value of
20 ppb
AOT30 seasonal sum of the difference
between an hourly concentration
at the threshold value of 30 ppb,
minus the threshold value of
30 ppb
AOT40
AOT60
AOTx
AP
A2p
APEX
APHEA(2)
APHENA
ApoB
ApoE
APX
aq
AQCD
AQI
AQS
AR
AR4
AR5
ARC
ARIC
ARIES
atm
ATP
ATPase
ATS
avg
AVHRR
AX
P
B
B1
seasonal sum of the difference
between an hourly concentration
at the threshold value of 40 ppb,
minus the threshold value of
40 ppb
seasonal sum of the difference
between an hourly concentration
at the threshold value of 60 ppb,
minus the threshold value of
60 ppb
family of cumulative, cutoff
concentration-based exposure
indices
activated protein
climate scenario in IPCC
(preliminary version of A2)
Air Pollutants Exposure (model)
Air Pollution on Health: a
European Approach (study)
Air Pollution and Health:
A European and North American
Approach
apolipoprotein B
apolipoprotein E
ascorbate peroxidase
aqueous form: (aq)O3
Air Quality Criteria Document
Air Quality Index
(U.S. EPA) Air Quality System
(database)
acoustic rhinometry
Fourth Assessment Report (AR4)
from the IPCC
Fifth Assessment Report (AR5)
from the IPCC
arginase variants (ex., ARG1,
ARG2, ARG1h4)
Atherosclerosis Risk in
Communities
(Atlanta) Aerosol Research and
Inhalation Epidemiology Study
atmosphere
adenosine triphosphate
adenosine triphosphatase;
adenosine triphosphate synthase
American Thoracic Society
average
advanced very high resolution
radiometer
Annex
beta, beta coefficient; regression
coefficient; standardized
coefficient; shape parameter;
scale parameter
boron
climate scenario in IPCC
XXXV
-------
B6
BAL
BALB/c
BALF
bb
BB
BC
B cells
B6C3F1
BDNF
BEAS-2B
BEIS
BELD
BIPM
BM
BMI
BMP
BP
BPD
bpm
Br
BRFSS
BS
BSA
Bsp, BSP
Bt, BT, bt
BTEX
BW
C
C3
C3
C4
C16:0
018:1
Ca
Ca
[Ca]
Ca2+
CA
CAA
mouse strain (C57BL/6J)
broncho-alveolar lavage
mouse strain
bronchoalveolar lavage fluid
bronchials
bronchial airways
black carbon
bone-marrow-derived
lymphocytes; B lymphocytes
mouse strain
brain-derived neurotrophic factor
human bronchial epithelial cell line
Biogenic Emissions Inventory
System
Biogenic Emissions Landcover
Database
International Bureau of Weights
and Measures
basement membrane
body mass index
P -type natriuretic peptide
blood pressure
biparietal diameter
breaths per minute
bromine
Behavioral Risk Factor
Surveillance System
black smoke
bovine serum albumin;
body surface area
black smoke particles
Bacillus thuringiensis; bacterium
proteins used in pesticides (or
genetically engineered plants
produce Bt toxin)
family of compounds (benzene,
toluene, ethylbenzene, and xylene)
body weight
carbon; concentration;
(Vitamin C, ascorbate)
degrees Celsius
carbon-13 isotope
mouse strain (C3H/HEJ)
plants that use only the Calvin
cycle for fixing the carbon dioxide
from the air
plants that use the Hatch-Slack
cycle for fixing the carbon dioxide
from the air
palmitic acid (saturated fatty acid)
unsaturated fatty acid
calcium
ambient concentration
calcium concentration
calcium ion
Canada (ICD-10-CA)
Clean Air Act
CALINE4
CAM
CAMP
CAMx
CAN
CAP(s)
CAR
CASAC
CASTNET
CAT
CB
C57BL/6
C57BL/6J
CBSA
C/C
CCSP
CD
CD-1
CDC
CF
CF2
C-fibers
CFR
CGRP
CH3
CH4
C2H2
C2H4
C3H
C3H6
CHAD
CH3Br
CHs-CHO
CH3CI
CH3-CO
CHD
CHF
C2H5-H
C3H/HeJ
CH3I
California line source dispersion
model for predicting air pollutant
concentrations near roadways
plants that use crassulacean acid
metabolism for fixing the carbon
dioxide from the air
Childhood Asthma Management
Program
Comprehensive Air Quality Model,
with extensions
Canada
concentrated ambient particles
centriacinar region
Clean Air Scientific Advisory
Committee
Clean Air Status and Trends
Network
catalase
carbon black; CMAQ mechanisms
(ex., CB04, CB05, CB06)
mouse strain
mouse strain
core-based statistical area
carbon of total carbon
Clara cell secretory protein
cluster of differentiation (various
receptors on T-cells: CD8+, CD44,
etc.);
criteria document (see AQCD)
mouse strain
Centers for Disease Control and
Prevention
charcoal-filtered; carbon filtered air
twice-filtered air (particulate filter
and activated charcoal filter)
afferent, slow, unmylenated nerves
innervating the respiratory system
Code of Federal Regulations
calcitonin gene-related peptide
methyl group
methane
acetylene
ethylene
mouse strain (C3H/HEJ or
C3H/OuJ)
propylene
Consolidated Human Activity
Database
methyl bromide
acetaldehyde
methyl chloride
acetyl radical(s)
coronary heart disease
congestive heart failure
ethane
mouse strain
methyl iodide
XXXVI
-------
CHIP
CH302"
CH3OOH
CHS
Cl
Cl
cr
CI2
CLE
CLM
CINO2
cm
cm2
CM
CMAQ
CN
CNA
CMS
CO
CO2
COD
Col-0
COP
COPD
COX-2
C-R
CRA
CRP
CS
CSA
csb, Csb
CSF
CST
CSTR
CSV
CT
CTM(s)
cum avg
Effects of Elevated Carbon Dioxide
and Ozone on Potato Tuber
Quality in the European Multiple
Site Experiment
methyl peroxy (radical)
acetic acid; methyl hydroperoxide
Child Health Study
confidence interval(s)
airborne O3 concentration at
microenvironment j
chlorine
chlorine ion
chlorine gas
Current Legislation (climate
scenario in IPCC)
chemiluminescence method
nitryl chloride
centimeter(s)
square centimeters
Clinical Modification (ICD-9-CM)
Community Multi-scale Air Quality
modeling system
constant atmospheric nitrogen
deposition (in PnET-CN
ecosystem model)
continental North America
central nervous system
carbon monoxide; Cardiac output
carbon dioxide
coefficient of divergence;
coefficient of determination
(Arabidopsis ecotype) Columbia-0
Conference of Parties (to the
UNFCCC)
chronic obstructive pulmonary
disease
cyclooxygenase 2 enzyme
concentration-response
Centra di ricerca per la
cerealicoltura (CRA) [The Centre
for Cereal Research] - Unit 5:
The Research Unit for Cropping
Systems in Dry Environments in
Bari, Italy (water-stressed
conditions)
C-reactive protein
corticosteroid
cross-sectional area; combined
statistical area
cockayne syndrome (cb)
gene/protein group A
colony-stimulating factor
central standard time
continuous stirred tank reactor
comma-separated values (a
spreadsheet format)
computer tomography
chemical transport model(s)
cumulative average
CUOt
CV, C.V.
cv, c.v.
CVD
CXC
CXCR2
CXR
CyS
Cys-LT
cyt
A, 6
AFEV,
AVD
2-D
3-D
DAHPS
DBP
DC(s)
DDM
DEP(s)
df
DGGE
DHA
DHAR
DHBA
DLEM
dm3
DMA
DOAS
DOC
DR
dt
DTH
DU
DW
E
EBC
EC
ECE
ECG
The cumulative stomatal uptake of
O3, using a constant O3 uptake
rate threshold (t) of nmol/m2/sec
coefficient of variation
cultivar
cardiovascular disease
chemokine family of cytokines,
with highly conserved motifcys-
xxx-cys (CXC) amino acid group
CXC chemokine receptor 2
(CXCR2)
Chest (x-ray) radiograph(s)
protein cysteines
cysteinyl leukotrienes (LTC4,
LTD4, LTE4)
cytosolic-free
delta, difference; change
change in FEN/!
change in dead space volume of
the respiratory tract
two-dimensional
three-dimensional
3-deoxy-D-arabino-heptulosonat-
7-phosphate synthase
diastolic blood pressure
dendritic cell(s)
direct decoupled method
diesel exhaust particle(s)
degrees of freedom
denaturing gradient gel
electrophoresis
dehydroascorbate
dehydroascorbate reductase
2,3-dihydroxybenzoic acid
Dynamic Land Ecosystem Model
cubic decimeter(s)
deoxyribonucleic acid
differential optical absorption
spectroscopy
dissolved organic carbon
type of human leukocyte antigens
(HLA-DR)
Portion of time-period spent in
microenvironment j
delayed-type hypersensitivity
Dobson unit(s)
dry weight
embryonic day (ex., E15, E16,
etc); [Vitamin] E
exposure to pollutant of ambient
origin
exhaled breath condensate (fluid)
elemental carbon
endothelin converting enzyme(s)
[i.e., ECE-1]
electrocardiogram
XXXV11
-------
ECOPHYS
ED
EGEA
EGEA2
EHC-93
ELF
EMI
Ena
ENA-78
eNO
eNOS
ENVISAT
EOTCP
EP
EPA
EPIC
ER
ESA
ET
ET,
ET2
ETS
EU
EUS
O
0PSII-max
f
F
F344
F2a
FA
FACE
FACES
fe
FC
physiological process modeling to
predict the response of aspen
forest ecosystems (modeling
growth and environmental stress in
Populus)
emergency department; embryonic
day (ex., EDS, ED20)
(The) Epidemiology (study on)
Genetics and Environment of
Asthma, (adults and children with
asthma)
follow-up study on EGEA (adults
with asthma only)
ambient PM reference sample
(urban dust [air particles] collected
in Ottawa Canada)
extracellular lining fluid
(U.S. EPA) Exposure Model for
Individuals
exposure to pollutant of
nonambient origin
epithelial cell-derived neutrophil-
activating peptide 78
exhaled nitric oxide
endothelial nitric oxide synthase
(EAS) Earth Observation satellite
European Open Top Chamber
Programme
epithelial cells
U.S. Environmental Protection
Agency
European Prospective
Investigation into Cancer and
Nutrition
emergency room
European Space Agency
extrathoracic; endothelin
(i.e., ET-1)
anterior nasal passages within the
extrathoracic (ET) region
oral airway and posterior nasal
passages within the extrathoracic
(ET) region
environmental tobacco smoke
European Union
eastern U.S.
Phi; calculated efficiency
maximum photochemical effective
quantum yield of PSII
Fraction of the relevant time period
female
Fischer 344 (rat strain)
8-isoprostane (major F2
prostaglandin [8 iso-PGF2a])
filtered air
free-air-CO2 enrichment (system)
Fresno Asthmatic Children's
Environment Study
frequency of breathing
fibrocartilaginous coat
FEF
FEF25-75
FEFx
FEM
FeNO
FEV,
FHM
FIA
Fin,
F inf,/
FLAG
FLRT
r nose
Fo
FPM
FR
FRAP
FRC
FRM
FRT
FstO-i
FURT
FVC
Fv/Fm
FVI
Y
Y-TOH
g, kg, mg, |jg, ng, pg
g
GAM
GCLC
GCLM
G-CSF
GD
GEE
forced expiratory flow
forced expiratory flow between the
times at which 25% and 75% of
the vital capacity is reached
forced expiratory flow after (x)%
vital capacity (e.g., after 25, 50, or
75% vital capacity)
Federal equivalent method
exhaled nitric oxide fraction
forced expiratory volume in
1 second
(USDA Forest Service) Forest
Health Monitoring Program
(USDA Forest Service) Forest
Inventory and Analysis Program
infiltration factor
infiltration factor for indoor
environment (i)
Federal land managers' air quality
related values workgroup
fractional uptake efficiency of the
lower respiratory tract (LRT)
fractional uptake efficiency via
nasal absorption
fraction of time spent in outdoor
microenvironments
Forest Pest Management
Federal Register
ferric reducing ability of plasma
functional residual capacity
Federal reference method
fractional uptake efficiency of the
respiratory tract (RT)
flux cut off threshold
fractional uptake efficiency of the
upper respiratory tract (URT)
forced vital capacity
a ratio: a measure of the maximum
efficiency of Photosystem II
fruits and vegetables index
gamma
gamma-tocopherol
gram(s), kilogram(s), milligram(s),
microgram(s), nanogram(s),
picogram(s)
granulocyte; guanosine
gram(s); gaseous form: (g)O3
generalized additive model(s)
conductance through boundary
layer and stomata
(glutathione genetic variant)
glutamate-cysteine ligase catalytic
subunit
(glutathione genetic variant)
glutamate-cysteine ligase modifier
subunit
granulocyte colony-stimulating
factor (receptor)
gestational day
generalized estimating equations
XXXV111
-------
GEOS (NASA) Goddard Earth Observing
System model
GEOS5 GEOS version 5
GEOS-Chem GEOS-Chemistry (tropospheric
model)
GFAP glial fibrillary acidic protein
GH growth hormone
GHG greenhouse gas
GLM(s) generalized linear model(s)
GMAO (NASA) Global Modeling and
Assimilation Office
GM-CSF granulocyte macrophage colony-
stimulating factor
GOME (ESA) Global Ozone Monitoring
Experiment (spectrometer)
GOMOS Global Ozone Monitoring by
Occultation of Stars (ESA
ENVISAT spectrometer measuring
long-term trends in O3)
G6P glucose-6-phosphate
G6PD glucose-6-phosphate
dehydrogenase
GPP gross primary production
G-proteins GTPases
GPT gas phase titration
GR glutathione reductase
GSH glutathione; reduced glutathione
GSO37GSO32~ guanine sulfonates
GSR glutathione reductase
GSS glutathione synthetase
GSSG glutathione disulfide
GST glutathione S-transferase
GSTM1 glutathione S-transferase
polymorphism M1 genotypes
(GSTM1-null, -GSTM1-sufficient)
GSTP1 glutathione S-transferase
polymorphism P1 genotypes
GTP guanosine triphosphate
GTPases G-proteins/enzymes
GWP global warming potential
GxE gene-environmental interaction
h hour(s)
h/day hour(s) per day
H; H+; H- atomic hydrogen, hydrogen ion;
hydrogen radical
3H radiolabeled hydrogen; tritium
H2 molecular hydrogen
ha hectare
HA hyaluronic acid, hospital admission
HA(s) hospital admission(s)
Hb hemoglobin
HbA1c glycosylated hemoglobin (blood
test)
HC(s) hydrocarbon(s)
HCFC(s) hydrochlorofluorocarbon(s)
HCHO formaldehyde
H2CO formaldehyde
HCO-
HDM
2HDM
HDMA
3He
HeJ
HEPA
HERO
12-HETE
HF
HFCs
Hg
HHP-C9
HIST
HLA
HLA-DR
HMOX
HMOX-1
HNE
HN02
HN03
HN04
HO
HO-
HO-1
H02-
H03-
H20
H202
H30+
HOCH2OOH
HONO
H02N02
HOONO
HOX
hPa
HPLC
HPOT
HR
HRmax
HRP
HRV
HSC
hs-CRP
H2SO4
HSP
HSP70
formyl (radical)
house dust mite
2nd-highest daily maximum
house dust mite allergen
non-radioactive isotope of helium
O3-resistant C3H mouse strain
(C3H/HeJ)
high efficiency particle air (filter)
Health and Environmental
Research Online, NCEA Database
System
12-Hydroxyeicosatetraenoic acid
(HRV signal) high-frequency
power
hydrofluorocarbons
mercury
1-hydroxy-1-hydroperoxynonane
histamine
human leukocyte antigen
human leukocyte antigen receptor
genes
Heme oxygenase
heme-oxygenase-1
(polymorphism)
4-hydroxynonenal
nitrous acid
nitric acid
pernitric acid
hydroxyl; heme oxygenase
hydroxyl radical
heme oxygenase 1
hydroperoxyl; hydroperoxy radical;
protonated superoxide
protonated ozone radical
water
hydrogen peroxide
hydronium ion
hydroxymethylhydroperoxide
nitrous acid
peroxynitric acid
pernitrous acid
hydrogen radical(s)
hectopascal
high-pressure liquid
chromatography
13-hydroperoxide linolenic acid
heart rate, hazard ratio
maximum heart rate
horseradish peroxidase
heart rate variability
Houston Ship Channel (Texas)
high-sensitivity C-reactive protein
sulfuric acid
high speed pellet (after centrifuge
spin)
heat shock protein 70
XX XIX
-------
HSS
5-HT
hv
HVAC
Hz
I
IARC
IAS
IBM
1C
ICAM-1
ICARTT
I CAS
ICC
ICD
ICD-9
ICD-10
ICEM
ICNIRP
ICP Forests
ICU
ICVE
IDW
IFN
IFN-Y
ig
IgA
igE
IGF-1
igG
igM
IHD
IL
IL-1P
lie
i.m.
IMPACT
IMPROVE
high speed supernatant (after
centrifuge spin)
5-hydroxytryptamine
Energy per photon of
electromagnetic energy at
frequency v
heating, ventilation, and air
conditioning
hertz
iodine
International Agency for Research
on Cancer
interalveolar septum
individual-based model or
modeling
inspiratory capacity;
intracloud (lightning flash)
intercellular adhesion molecule 1
International Consortium for
Atmospheric Research on
Transport and Transformation
Inner City Asthma Study
intraclass correlation coefficient
implantable cardioverter
defibrillator(s); International
Classification of Diseases
International Classification of
Disease 9th revision
International Classification of
Disease 10th revision
Indoor Chemistry and Exposure
Model
International Commission on
Non-ionizing Radiation Protection
International Cooperative
Programme on Assessment of Air
Pollution Effects on Forests
Intensive Care Unit
ischemic cerebrovascular events
inverse-distance-weighted
interferon (e.g., IFN-Q)
interferon-gamma
immunoglobulin (e.g., IgE)
immunoglobulin A
immunoglobulin E
insulin-like growth factor 1
immunoglobulin G
immunoglobulin M
ischemic heart disease
interleukin (e.g., IL-2, IL-4, IL-6,
IL-8, etc.)
interleukin-1p
isoleucine
intramuscular (route)
Interactive Modeling Project for
Atmospheric Chemistry and
Transport
Interagency Monitoring of
Protected Visual Environment
IN intranasal
INF interferon
inh inhalation
iNKT invariant (type I) natural killer
T-cell
iNOS inducible nitric oxide synthase
INRA National agronomical research
institute (INRA) in Thiverval-
Grignon. France (adequately-
watered conditions)
INTRASTAND a stand-level model designed for
hourly, daily and annual integration
of forest carbon and water cycle
fluxes
I/O indoor-outdoor ratio
IOM Institute of Medicine
i.p. intraperitoneal (route)
IPCC Intergovernmental Panel on
Climate Change
IPCC-A2 Intergovernmental Panel on
Climate Change 2nd Assessment
Report
IPCC-AR4 Intergovernmental Panel on
Climate Change 4th Assessment
Report
IPCC-AR5 Intergovernmental Panel on
Climate Change 5th Assessment
Report
IPCC-TAR Intergovernmental Panel on
Climate Change Third Assessment
Report
IPMMI International Photolysis Frequency
Measurement and Modeling Inter-
comparison
IQR interquartile range
IR infrared
I/R ischemia-reperfusion
IRIS Integrated Risk Information
System
IRP Integrated Review Plan for the
Ozone National Ambient Air
Quality Standards
ISA Integrated Science Assessment
ISCCP International Satellite Cloud
Climatology Project
ISO International Standards
Organization
8-iso-PGF 8-isoprostane
IT intratracheal
IU International Units
IUGR intrauterine growth restriction
i.v. intravenous (route)
IVF in vitro fertilization
j Microenvironment
JA jasmonic acid
Jmax maximum rate of electron transport
(for regeneration of RuBP)
JNK jun N-terminal kinase
JPL Jet Propulsion Laboratory
K kappa
xl
-------
KB kappa B
k dissociation rate; root:shoot
allometric coefficient; rate of O3
loss in the microenvironment
K potassium
K+ potassium ion
Ka intrinsic mass transfer
coefficient/parameter
KC keratinocyte-derived chemokine
kg kilogram
Kg mass transfer coefficient for gas
phase
kHz kilohertz
kJ kilojoules
Kl mass transfer coefficient for liquid
phase
km kilometer
KM particle optical reflectance
KML keyhole markup language
KMZ zipped KML computer language
KO knockout
Kr reaction rate constant
KROFEX Krauzberg Ozone Fumigation
Experiment
L, dL, ml, |jL Liter, deciliter, milliLiter, microLiter
LO Lag (e.x., Lag 0, Lag 1, etc.)
LAI leaf area index
LBL Lawrence Berkeley Laboratory
LBLX Lawrence Berkeley Laboratory
model including airflow from
natural ventilation
Lb(s) pound(s)
LBW low birth weight
LC50 median lethal concentration
LCL lower 95th% confidence limit
LDH lactate dehydrogenase
LDL low-density lipoprotein; lower
detectable level
LF (HRV signal) low-frequency power
LFHFR low frequency/high frequency
(ratio)
LFT lower free troposphere
LI labeling index
LIDAR Light Detection and Ranging
(remote sensing system)
LIF laser-induced fluorescence
LINKAGES individual-based model of forest
succession
LIS lateral intercellular space
LLJ low-level jet
L/min liters per minute
Ln Natural logarithm
LnRMSSD natural log of RMSSD; measure of
HRV
InSDNN natural log of the standard
deviation of NN intervals in
anEKG
LOAEL
LOD
LOEL
LOESS
LOP
LOSU
LOWESS
LOX-1
LPS
LRS
LRT
LSI
LT
LT-a
LTA
LUR
LVEDD
LVEDP
LWC
M
|jg/m3
|jm
m, cm, |jm, nm
M
M, mM, |jM, nM, pM
m2
m3
M#
M2
M7
M12
ma
mAOT
MAP
MAPK
MAQSIP
MARAT
MARCO
lowest observed adverse effect
level
limit of detection
lowest-observed-effect level
locally weighted scatterplot
smoothing
lipid ozonation products
level of scientific understanding
locally weighted scatter plot
smoother
Lipoxygenase; lectin-like oxidized
low density lipoprotein receptor-1
lipopolysaccharide
lower respiratory symptoms
lower respiratory tract; lower
airways; Long range transport
local standard time
leukotriene (e.g., LTB4, LTC 4,
LTD4, LTE4); local time
lymphotoxin-a
lymphotoxin-alpha
land use regression
left ventricular chamber
dimensions at end diastole
left ventricular end diastolic
pressure
liquid water content
mu, micro
microequivalent
microgram
micrograms per cubic meter
micrometer, micron
meter(s), centimeter(s),
micrometer/[micron](s),
nanometer(s)
male
Molar, milNMolar, microMolar,
nanoMolar, picoMolar
square meters
cubic meters
Month (M1 Monthl; M2 Month2;
M3 MonthS; M4 Month4)
type of muscarinic receptor
7-hour seasonal mean
12-hour seasonal mean of O3
moving average
modified accumulated exposure
over threshold
mitogen-activated protein; mean
arterial pressure
mitogen-activated protein
kinase(s), MAP kinase
Multiscale Air Quality Simulation
Platform (model)
Mid-Atlantic Regional Assessment
Team
Macrophage receptor with
collagenous structure
xli
-------
max
MBL
MCA
MGCP
Mch; MCh
MCM
MCP-1
MDA
MDAR
MDI
MDL
MED
MEF50%
MEGAN
MeJA
MENTOR
METs
MFR
Mg
MGDG
mg/m3
MHC
mi
Ml
MIESR
min
MIP
MIP-2
ml
mL/min
MLN
Mm
mm
MM Mt.
MM5
MMAD
MMEF
mmHg
MMMD
MMP-2
MMP-3
MMP-9
MMSP
Mn
maximum
marine boundary layer
minimum cross-sectional area
Mountain Cloud Chemistry
Program
methacholine
master chemical mechanism
monocyte chemotactic protein 1
malondialdehyde
monodehydroascorbate reductase
Mediterranean diet index
minimum detection level
minimal erythema dose
maximal midexpiratory flow at 50%
of forced vital capacity
model of emissions of gases and
aerosols from nature
methyl jasmonate
Modeling Environment for Total
Risk Studies
metabolic equivalent unit(s)
[of work]
Maximum Feasible Reduction
magnesium
monogalactosyl diacylglycerol
milligrams per cubic meter
major histocompatibility complex
mile(s)
myocardial infarction, "heart
attack"
matrix isolation electron spin
resonance (spectroscopy)
minute; minimum
macrophage inflammatory protein
macrophage inflammatory
protein 2
milliliter
milliliter(s) per minute
mediastinal lymph node
megameter
millimeter(s)
Mt. Mitchell site
National Center for Atmospheric
Research/Penn State Mesoscale
Model (version 5)
mass median aerodynamic
diameter; mass median
aerodynamic density
maximal midexpiratory flow
millimeters of mercury
mean maximum mixing height
depth
matrix metalloproteinase-2
matrix metalloproteinase-3
metalloproteinase-9
Mount Mitchell State Park, NC
manganese
M/N
MnSOD
mo
MOA(s)
MOBILE
MOBILE6
MODNR
MONICA
MoOx
MOSES
MOVES
MOZAIC
MOZART
MPAN
MPO
MQL
MRI
mRNA
ms
MS
MSA
MSL
MS/MS
MT
MT, Mt
MT1
MTBE
mtDNA
Mtn
MW
MyD88
n, N
N
15N
N2
Na
NA
pooled data from mouth and nasal
exposure
Manganese superoxide dismutase
month(s)
mode(s) of action
(U.S. EPA) mobile vehicle
emission factor model (on-road
vehicles)
vehicle emissions modeling
software version 6; replaced by
MOVES
Missouri Department of Natural
Resources
Monitoring of Trends and
Determinants in Cardiovascular
Disease
molybdenum oxides
Met Office Surface Exchange
Scheme
Motor Vehicle Emission Simulator
(replaced MOBILE6; for estimating
emissions from cars, trucks, and
motorcycles
Measurement of Ozone and Water
Vapor by Airbus In-Service Aircraft
Model for Ozone and Related
chemical Tracers
peroxymethacryloyl nitrate;
peroxy-methacrylic nitric anhydride
myeloperoxidase
Minimum quantification limit
magnetic resonance imaging;
Midwest Research Institute;
Meteorological Research Institute
messenger RNA
millisecond(s)
mass spectrometry; Mt.
Moosilauke site
Metropolitan Statistical Area;
methane sulfonic acid
mean sea level
tandem mass spectrometry
million ton(s); metric ton(s)
metallothionein
mitochondria
methyl-tertiary-butyl ether
mitochondrial DNA
mountain
molecular weight
myeloid differentiation primary
response gene 88
number; number of observations
nitrogen; North; nasal exposure by
natural breathing
nitrogen-15, stable isotope of
nitrogen
molecular nitrogen; nonreactive
nitrogen
sodium
noradrenaline; North American
xlii
-------
NA; N/A not available; not applicable
Na+ sodium ion
NAAQS National Ambient Air Quality
Standards
NAD nicotinamide adenine nucleotide
NADH reduced nicotinamide adenine
dinucleotide; nicotinamide adenine
dinucleotide dehydrogenase
NADP National Atmospheric Deposition
Program
NADPH reduced nicotinamide adenine
dinucleotide phosphate
NADPH-CR reduced nicotinamide adenine
dinucleotide phosphate -
cytochrome c reductase
NaE sodium erythorbate
NAG N-acetyl-glucosaminidase
Na-K-ATPase sodium-potassium-dependent
adenosine triphosphatase
NAMS National Ambient Monitoring
Stations
NAPAP National Acid Precipitation
Assessment Program
NAPBN National Air Pollution Background
Network
NARE North Atlantic Regional
Experiment
NARSTO North American Regional Strategy
for Tropospheric Ozone
NAS National Academy of Sciences;
Normative Aging Study
NASA National Aeronautics and Space
Administration
NBS National Bureau of Standards
NBTH 3-methyl-2-benzothiazolinone
acetone azine
NCE net carbon exchange
NCEA National Center for Environmental
Assessment
NCEA-RTP NCEA Division in Research
Triangle Park, NC
NCHS National Center for Health
Statistics
NCICAS National Cooperative Inner-City
Asthma Study
NCLAN National Crop Loss Assessment
Network
NCore National Core multipollutant
monitoring network
NC-R resistant clones of white clover
NC-S sensitive clones of white clover
ND; n.d. not detectable; not detected;
no data
2ndHDM 2nd-highest daily maximum
NDF neutral detergent fiber
NEE net ecosystem exchange
(of carbon or CO2)
NEI National Emissions Inventory
NEM National Ambient Air Quality
Standards Exposure Model
NEP Net Ecosystem Production
NERL National Exposure Research
Laboratory
NESCAUM Northeast States for Coordinated
Air Use Management
NF National Forest; non-filtered air
NF-KB nuclear factor kappa B
ng nanogram(s)
NGF nerve growth factor
NH northern hemisphere
NH3 ammonia
NH4+ ammonium ion
NH4HSO4 ammonium bisulfate
(NH4)2HSO4 ammonium sulfate
NHANES National Health and Nutrition
Examination Survey
NHANES III National Health and Nutrition
Examination Survey III
NHAPS National Human Activity Pattern
Survey
NHEERL (U.S. EPA) National Health and
Environmental Effects Research
Laboratory
NHIS National Health Interview Survey
(NH4)2SO4 ammonium sulfate
NIH National Institutes of Health
NIST National Institute of Standards and
Technology
NK natural killer cells; neurokinin
NKT natural killer T-cells
NL nasal lavage
NLF nasal lavage fluid
NM National Monument
NMHC(s) nonmethane hydrocarbon(s)
NMMAPS National Morbidity, Mortality, and
Air Pollution Study
NMOC(s) nonmethane organic compound(s)
NMVOCs nonmethane volatile organic
compounds
NN normal-to-normal (NN or RR) time
interval between each QRS
complex in the EKG
NNK 4-(N-nitrosomethylamino)-1-
(3-pyridyl)-1 -butanone
nNOS neuronal nitric oxide synthase
(NOS)
NO nitric oxide
• NO nitric oxide concentration
(interpunct NO)
NO2 nitrogen dioxide
NO3; NO3- nitrate, nitrate radical
NO3~ nitrate, nitrate ion
N2O nitrous oxide
N2O5 dinitrogen pentoxide
NOAA National Oceanic and Atmospheric
Administration
NOAEL no observed adverse effect level
xliii
-------
NOS
NOX
NOY
NOZ
NP
NPP
NPS
NQO1
NQOIwt
NR
Nr
NRC
Nrf-2
Nrf2-ARE
NS; n.s.
NSAID
NSBR
NSF
NTE
NTN
NTP
NTRMs
NTS
NWR
NWS
NZW
O
02-
102
03
1803
03*
OAQPS
OAR
OEMs
OC
nitric oxide synthase (types,
NOS-1, NOS-2, NOS-3)
nitrogen oxides, oxides of nitrogen
(N0 + N02)
sum of NOX and NOZ; odd
nitrogen species; total oxidized
nitrogen
sum of all inorganic and organic
reaction products of NOX (MONO,
HNO3, HNO4, organic nitrates,
particulate nitrate, nitro-PAHs,
etc.)
National Park
net primary production
National Park Service, U.S.
Department of the Interior
NAD(P)H-quinone oxidoreductase
(genotype)
NAD(P)H-quinone oxidoreductase
wild type (genotype)
not reported
reactive nitrogen
National Research Council
nuclear factor erythroid 2-related
factor 2
NF-e2-related factor 2-antioxidant
response element
nonsignificant; non-smoker;
national seashore; natural spline
non-steroidal anti-inflammatory
agent
nonspecific bronchial
responsiveness
National Science Foundation
nasal turbinate epithelial (cells)
National Trends Network
National Toxicology Program
NIST Traceable Reference
Materials
nucleus of the solitary tract (in
brainstem)
national wildlife refuge
National Weather Service
New Zealand white (rabbit)
oxygen; horizon forest floor
oxygen-18, stable isotope of
oxygen
molecular oxygen
superoxide
superoxide radical
singlet oxygen
ozone
(oxygen-18 labeled) ozone
electronically excited ozone
Office of Air Quality Planning and
Standards
Office of Air and Radiation
observationally based methods
organic carbon
OD
0(1D)
OH, OH-
8-OHdG
OLS
OMI
ON
ONOO"
0(3P)
OPE
OPECs
OR
ORD
OSHA
OTC
OuJ
OVA
OX
OxComp
oz
P
P
P450
p53
P90
PACF
PAD
PAF
PAH(s)
PAI-1
PAL
PAMS
PAN
Pa02
PAPA
PAR
Patm
p-ATP
Pb
PEL
outer diameter; optical density
electronically excited oxygen atom
hydroxyl group, hydroxyl radical
8-hydroxy-2'-deoxyguanosine
ordinary least squares
Ozone Monitoring Instrument
Ontario
peroxynitrate ion
ground-state oxygen atom
ozone production efficiency
Outdoor Plant Environment
Chambers
odds ratio
Office of Research and
Development
Occupational Safety and Health
Administration
open-top chamber
O3-sensitive C3H mouse strain
(C3H/OuJ)
ovalbumin
odd oxygen species; total oxidants
oxidative capacity of the
atmosphere
ounce(s)
pressure in atmospheres; plants
grown in pots; phosphorus;
penetration fraction of O3 into the
microenvironment; pulmonary
region
probability value
cytochrome P450
cell cycle protein gene
90th percentile of the absolute
difference in concentrations
partial autocorrelation function of
the model residuals
peripheral arterial disease;
pollutant-applied dose
platelet-activating factor;
paroxysmal atrial fibrillation
polycyclic aromatic hydrocarbon(s)
plasminogen activator fibrinogen
inhibitor-1
phenylalanine ammonia lyase
Photochemical Assessment
Monitoring Stations network
peroxyacetyl nitrate
arterial oxygen pressure
Public Health and Air Pollution in
Asia
photosynthetically active radiation;
proximal alveolar region
Pressure in atmospheres
para-acetam idophenol
Lead
planetary boundary layer;
peripheral blood lymphocytes
xliv
-------
PBM
PEN
PBPK
PBS
PC
PC2o
PC5
PCA
PC-ALF
PCD
PCI
pCNEM
PCO2
pCO2
PCR
PCR-DGGE
PD
PD2o
PD10oSRaw
PDI
PE
PEF
PEFR
PEFT
PEG-CAT
PEG-SOD
PEM(s)
Penh
PEPc
PFD
PFT
pg
PG
population-based model or
modeling
C-phenyl N-tert-butyl nitrone
physiologically based
pharmacokinetic (model)
phosphate buffered saline
phosphatidylchloline
provocative concentration that
produces a 20% decrease in
forced expiratory volume in
1 second
provovative concentration that
produces a 20% decrease in FEN/!
provocative concentration that
produces a 50% decrease in
forced expiratory volume in
1 second
principal component analysis
1 -palmitoyl-2-(9-oxonononoyl)-sn-
glycero-3-phosphocholine
programmed cell death
picryl chloride
Canadian version of National
Ambient Air Quality Standards
Exposure Model
Average partial pressure of O2 in
lung capillaries
partial pressure of carbon dioxide
polymerase chain reaction
PCR-denaturing gradient gel
electrophoresis
pregnancy day
provocative dose that produces a
20% decrease in
provocative dose that produces a
20% decrease in
provocative dose that produces a
100% increase in sRAW
provocative dose that produces a
100% increase in SRaw
pain on deep inspiration
postexposure,
phosphatidylethanolamine
peak expiratory flow
peak expiratory flow in
0.75 second
peak expiratory flow rate
time to peak flow
polyethylene glycol-catalase
polyethylene glycol-superoxide
dismutase
personal exposure monitor(s)
enhanced pause
phosphoenolpyruvate carboxylase
photosynthetic flux density
pulmonary function test
picogram(s)
prostaglandin (e.g., PGE2, PGF2);
phosphatidylglycerol
6PGD
PGE2
PGF2a
PGHS-2
PGP
PGSM
PH
PHA
PI
PIF
PiZZ
PK
pKa
PLFA
PM
PMX
PM2
6-phosphogluconate
dehydrogenase
prostaglandin E2
prostaglandin F2-alpha
prostaglandin endoperoxide G/H
synthase 2
protein gene product
(e.g., PGP9.5)
Plant Growth Stress Model
relative acidity; Log of the
reciprocal of the hydrogen ion
concentration
phytohemagglutinin A
phosphatidylinositol; probability
interval; posterior interval
peak inspiratory flow
respiratory phenotype
pharmaco kinetics
dissociation constant
phospholipid fatty acid
particulate matter
Particulate matter of a specific size
range not defined for regulatory
use. Usually X refers to the 50%
cut point, the aerodynamic
diameter at which the sampler
collects 50% of the particles and
rejects 50% of the particles.
The collection efficiency, given by
a penetration curve, increases for
particles with smaller diameters
and decreases for particles with
larger diameters. The definition of
PMX is sometimes abbreviated as
"particles with a nominal
aerodynamic diameter less than or
equal to X |jm" although X is
usually a 50% cut point.
In general terms, particulate matter
with an aerodynamic diameter less
than or equal to a nominal 2.5 |jm;
a measurement of fine particles. In
regulatory terms, particles with an
upper 50% cut-point of 2.5 |jm
aerodynamic diameter (the 50%
cut point diameter is the diameter
at which the sampler collects 50%
of the particles and rejects 50% of
the particles) and a penetration
curve as measured by a reference
method based on Appendix L of 40
CFR Part 50 and designated in
accordance with 40 CFR Part 53,
by an equivalent method
designated in accordance with 40
CFR Part 53, or by an approved
regional method designated in
accordance with Appendix C of 40
CFR Part 58.
xlv
-------
PM10
PM10
PM10
pSSMAPK
PM-CAMx
PMN(s)
PMT
PND
pNEM
PnET
PNN
In general terms, particulate matter
with an aerodynamic diameter less
than or equal to a nominal 10 |jm;
a measurement of thoracic
particles (i.e., that subset of
inhalable particles thought small
enough to penetrate beyond the
larynx into the thoracic region of
the respiratory tract). In regulatory
terms, particles with an upper 50%
cut-point of 10 ± 0.5 |jm
aerodynamic diameter (the 50%
cut point diameter is the diameter
at which the sampler collects 50%
of the particles and rejects 50% of
the particles) and a penetration
curve as measured by a reference
method based on Appendix J of 40
CFR Part 50 and designated in
accordance with 40 CFR Part 53
or by an equivalent method
designated in accordance with 40
CFR Part 53.
In general terms, particulate matter
with an aerodynamic diameter less
than or equal to a nominal 10 |jm
and greater than a nominal 2.5
|jm; a measurement of thoracic
coarse particulate matter or the
coarse fraction of PMi0. In
regulatory terms, particles with an
upper 50% cut-point of 10 |jm
aerodynamic diameter and a lower
50% cut-point of 2.5 |jm
aerodynamic diameter (the 50%
cut point diameter is the diameter
at which the sampler collects 50%
of the particles and rejects 50% of
the particles) as measured by a
reference method based on
Appendix O of 40 CFR Part 50 and
designated in accordance with 40
CFR Part 53 or by an equivalent
method designated in accordance
with 40 CFR Part 53.
The PM10-2.5 concentration of
PMio-2.5 measured by the 40 CFR
Part 50 Appendix O reference
method which consists of currently
operated, co-located low-volume
(16.7 Lpm) PM10 and PM2.5
reference method samplers.
p38 mitogen-activated protein
kinase(s)
Comprehensive Air Quality Model
with extensions and with
particulate matter chemistry
polymorphonuclear leukocyte(s)
photomultiplier tube
post natal day
probabilistic National Exposure
Model
Photosynthetic EvapoTranspiration
model
proportion of interval differences of
successive normal-beat intervals
in EKG
PNN50
P02
POC
POD
polyADPR
POMS
ppb
ppb-h
ppbv
pphm
ppm
ppm-h
ppmv
PPN
PPPs
ppt
pptv
PQH2
PR
PR-1
PRB
preproET-1
PRYL
PS
PS
PS II
PSA
PSC
PTB
PTR-MS
PU,PUL
PUFA(s)
PV
PVCD
PVD
PVOCs
PWM
PWTES
Pxase
QA
QC
proportion of interval differences of
successive normal-beat intervals
greater than 50 ms in EKG
partial pressure of oxygen
particulate organic carbon
peroxidase
poly(adenosinediphosphate-
ribose)
Portable Ozone Monitoring
Systems
parts per billion
parts per billion per hour
parts per billion by volume
parts per hundred million
parts per million
parts per million hours; weighted
concentration values based on
hourly concentrations: usually
summed over a certain number of
hours, day(s), months, and/or
season.
parts per million by volume
peroxypropionyl nitrate;
peroxypropionic nitric anhydride
power plant plumes
parts per trillion
parts per trillion by volume
plastoquinone
pathogenesis-related (protein)
promoter region 1
policy-relevant background
pre-protein form of ET-1 mRNA
predicted relative yield (biomass)
loss
penalized spline
paradoxical sleep
Photosystem II: enzyme that uses
light to obtain electrons from water
(for photosynthesis).
picryl sulfonic acid
polar stratospheric clouds
preterm birth
proton-transfer-reaction mass
spectroscopy
pulmonary
polyunsaturated fatty acid(s)
potential vorticity
peripheral vascular and
cerebrovascular disease
peripheral vascular disease
photochemical volatile organic
compounds
pokeweed mitogen
(left ventricular) posterior wall
thickness at end systole
peroxidase
Quality Assurance
quality control
xlvi
-------
QCE
qNP
C|NP
qP
QRS
QT
QTc
r
R, r
r2
R2
R2,r2
RACM
RADM
rALP
RAMS
RANTES
Raw
RB
RBC(s)
rbcL
rbcS
R'CO acyl
R'C(0)-02
rcdl
RCD3
RCP
RDBMS
Re
REHEX
RER
RF
RGR
RH
RIOPA
RL
RLKs
RMNP
RMR
rMSSD
quasi continuous exercise
non-photochemical quenching
non-photochemical quenching
photochemical quenching
A complex of three distinct
electrocardiogram waves which
represent the beginning of
ventricular contraction
interval measure of the time
interval between the start of the Q
wave and the end of the T wave in
the heart's electrical cycle
corrected QT interval
Pearson correlation coefficient
correlation coefficient
correlation coefficient
multiple regression correlation
coefficient
coefficient of determination
Regional Atmospheric Chemistry
Mechanism
Regional Acid Deposition Model
recombinant antileukoprotease
Regional Atmospheric Modeling
System
regulated upon activation, normal
T-cell expressed and secreted
(cells)
airway resistance
respiratory bronchiole
red blood cell(s); erythrocyte(s)
Rubisco large subunit
Rubisco small subunit
acyl carrier protein
acyl peroxy
Arabidopsis mutant radical
induced cell death
rod-cone dysplasia 3
Representative Concentration
Pathways
Relational Database Management
Systems
Reynolds number
Regional Human Exposure Model
rough endoplasmic reticulum;
Respiratory exchange ratio
radiative forcing
relative growth rate
relative humidity
Relationship of Indoor, Outdoor,
and Personal Air (study)
total pulmonary resistance
receptor-like/Pelle kinase group
Rocky Mountain National Park,
Colorado
resting metabolic rate
root mean squared differences
between adjacent normal-to-
normal heartbeat intervals
Rn
RNA
RO2
ROG
ROI
RONO2
ROOM
ROONO2, RO2NO2
ROS
RPD
RR
RRMS
RT
RT
RTLF
RuBisCO; Rubisco
RuBP
a
s
S
s.c.
SA
SAB
SAC
SAG21
SAI
S-allele
SAMD
SaO2
SAPALDIA
SAPRC
SAR
SAROAD
SAWgrp
SBNF
SBP
SBUV
SC
Sc
nasal resistance
ribonucleic acid
organic peroxyl; organic peroxy
reactive organic gases
reactive oxygen
intermediate/superoxide anion
organic nitrate
organic peroxides
peroxy nitrate
reactive oxygen species
relative percent difference
normal-to-normal (NN or RR) time
interval between each QRS
complex in the EKG; risk ratio;
relative risk; respiratory rate
relatively remote monitoring sites
respiratory tract
transepithelial resistance
respiratory tract lining fluid
ribulose-1,5-bisphosphate
carboxylase/oxygenase
ribulose bisphosphate
sigma, standard deviation
sigma-g; (geometric standard
deviation)
second
Short; smoker; sulfur; South
subcutaneous (route)
salicylic acid
Science Advisory Board
Staphylococcus aureus Cowan
1 strain
senescence
Systems Applications International
short-allele
S-adenosyl methionine
decarboxylase
oxygen saturation of arterial blood
Study of Air Pollution and Lung
Diseases in Adults
Stratospheric Processes and their
Role in Climate; Statewide Air
Pollution Research Center,
University of California, Riverside
systemic acquired resistance
Storage and Retrieval of
Aerometric Data (U.S. EPA
centralized database; superseded
by Aerometric Information
Retrieval System [AIRS])
small airway function group
San Bernardino National Forest,
California
systolic blood pressure
Solar Backscatter Ultraviolet
Spectrometer
stratum corneum
scandium
xlvii
-------
SCAQS
SCE(s)
SD
SDNN
SE
SEBAS
sec
Sess.
SEM
SENP
SES
SF
SF6
SGA
sRaw
SH
SHEDS
SHEN
slCAM-1
SIDS
SIGMOID
SINIC
SIP
SIPK
SK
SLA
SLAC1
SLAMS
SM
SMD
SME
SMNP
SMOKE
SN
SNAAQS
SNP(s)
S02
S042"
SOC
Southern California Air Quality
Study
sister chromatid exchange(s)
standard deviation;
Sprague-Dawley rat
standard deviation normal-to-
normal (NN or RR) time interval
between each QRS complex in the
EKG
standard error
Social Environment and
Biomarkers of Aging Study
second
session
simultaneously extracted metal;
standard error of the mean;
scanning electron microscopy
Sequoia National Park, California
socioeconomic status
San Francisco Bay Area
sulfur hexafluoride (tracer gas)
small for gestational age
specific airway conductance
Shenandoah National Park site
Stochastic Human Exposure and
Dose Simulation
Shenandoah National Park
soluble intercellular adhesion
molecule
sudden infant death syndrome
sigmoid weighted summed
concentration
Simple Nitrogen Cycle model
State Implementation Plan
salicylic acid (SA) induced protein
kinase
shikimate kinase
specific leaf area
(protein) slow anion channel
associated 1
State and Local Air Monitoring
Stations
smooth muscle
soil moisture deficit
soybean oil methyl ester
Great Smoky Mountain National
Park (North Carolina and
Tennessee)
Spare-Matrix Operator Kernel
Emissions
normalized slope of the alveolar
plateau
Secondary National Ambient Air
Quality Standards
single-nucleotide polymorphism
sulfur dioxide
sulfate
soil organic carbon
SOCS Salmeterol Off Corticosteroids
Study
SOD superoxide dismutase
SOS Southern Oxidant Study
SOX sulfur oxides
SoyFACE Soybean Free Air gas
Concentration Enrichment
(Facility)
SP surfactant protein (e.g., SPA,
SPD); substance P
SP-A surfactant protein-A
SPF specific pathogen free
SPMs special purpose monitors
SP-NK substance P - neurokinin receptor
complex
sRaw, specific airway resistance
SRBC sheep red blood cell
SRES Special Report on Emissions
Scenarios
SRM standard reference method
SRP standard reference photometers
SSCP single-strand conformation
polymorphism
12951/SvlmJ mouse strain
STE stratosphere-troposphere
exchange
STEP Stratospheric-
Tropospheric-exchange Project
STN speciation trends network
sTNFRI soluble tumor necrosis factor
receptor 1
STP standard temperature and
pressure
STPD standard temperature and
pressure, dry
STRF Spatio-Temporal Random Field
(theory)
subscript i Index of indoor microenvironments
subscript o Index of outdoor
microenvironments
subscript o,i Index of outdoor
microenvironments adjacent to a
given indoor microenvironment /
SUMOO sum of all hourly average
concentrations
SUM06 seasonal sum of all hourly average
concentrations > 0.06 ppm
SUM07 seasonal sum of all hourly average
concentrations > 0.07 ppm
SUM08 seasonal sum of all hourly average
concentrations a 0.08 ppm
SURE Sulfate Regional Experiment
Program
SVE supraventricular ectopy
S-W square-wave
SWS slow wave sleep
SZA solar zenith angle
T tau, photochemical lifetime;
atmospheric lifetime
xlviii
-------
t t-test statistical value; t statistic
T time; duration of exposure
T-cell(s) T lymphocyte(s), thymus-
dependent lymphocytes
T1 first trimester
T2 second trimester
T3 triiodothyronine
T3 third trimester
T4 thyroxine
TAR IPCC Third Assessment Report
TAR WGI IPCC Third Assessment Report of
Working Group I
TB tracheobronchial; terminal
bronchioles; tuberculosis
TBA thiobarbituric acid
TEARS thiobarbituric acid reactive
substances
TC total carbon
99mTc Technetium-99m
T-cells T-lymphocytes, Thymus-derived
lymphocytes
99mTc-DTPA 99mTc-
diethylenetriaminepentaacetic acid
Tco core temperature
TOLAS Tunable Diode Laser Absorption
Spectrometer
Te expiratory time
TEM transmission electron microscopy;
Terrestrial Ecosystem Model
TES Tropospheric Emission
Spectrometer
TexAQS Texas Air Quality Field Study
Tg teragram(s)
TGF transforming growth factor
TGF p transforming growth factor beta
Th T helper cell type
Th2 T helper cell type 2
THC Total hydrocarbon content
tHcy total homocysteine
Ti inspiratory time
Ti titanium
TIA transient ischemic attack
TIMP-2 tissue inhibitor of matrix
metalloprotease-2
TiO2 titanium dioxide
TLC total lung capacity
TLNISE two-level normal independent
sampling estimation
Tlr Toll-like receptor gene
TLR Toll-like receptor protein (ex.,
TLR2, TLR4)
TMPO tetramethylphrrolise 1-oxide
TNC total nonstructural carbohydrate
TNF tumor necrosis factor (e.g., TNF-a)
TNF-308 tumor necrosis factor genotype
TNF-a tumor necrosis factor alpha
TNFR tumor necrosis factor receptor
TOMS
TOPSE
(PA
TPLIF
TRAMP
TREGRO
TRIFFID
TRIM
TRIM.Expo
TRP
TSH
TSP
TTFMS
TWA
TX
TXB2
UA
UAM
UCL
UDGT
UDP
U.K.
UNECE
UNEP
UNFCCC
U-O
U-O2"
U-03"
URI
URS
URT
U.S.
USC; U.S.C.
USDA
USFS
USGCRP
USGS
UV
UV-A
Total Ozone Mapping/Monitoring
Satellite; total ozone mapping
spectrometer
Tropospheric Ozone Production
About the Spring Equinox
tissue plasminogen activator
two-photon laser-induced
fluorescence
TexAQS-ll Radical and Aerosol
Measurement Project
Tree Growth Model
Top-down Representation of
Interactive Foliage and Flora
Including Dynamics
Total Risk Integrated Methodology
(model)
Total Risk Integrated Methodology
Exposure Event (model)
transient receptor potential (ion
channel[s], ex., TRP-A1, TRP-V1,
TRP-M8)
thyroid stimulating hormone
total suspended particles
two-tone frequency-modulated
spectroscopy
time-weighted average
thromboxane (e.g., TXB2)
thromboxane B2
uric acid; urate
Urban Airshed Model
upper 95th% confidence limit
UDP -galactose-1,2,-diacylglycerol
galactosyltransferase
uridine diphosphate
United Kingdom
United Nations Economic
Commission for Europe
United Nations Environmental
Programme
United Nations Framework
Convention on Climate Change
epioxides formed from uric acid
peroxides formed from uric acid
ozonides formed from uric acid
upper respiratory infection
upper respiratory symptoms
upper respiratory tract; upper
airways
United States (of America)
U.S. Code
U.S. Department of Agriculture
U.S. Forest Service
U.S. Global Change Research
Program
U.S. Geological Survey
ultraviolet radiation
ultraviolet radiation at wavelengths
of 320 to 400 nm
xlix
-------
UV-B
UV-C
UV-DIAL
V
V, mV, |JV
VA
Val
VC
VCAM
Vd
VD
VE
VEGF
VEmax
Vmax
Vmax25%
Vmax 50%
Vmax75o/0
VMD
Vn
VO2
VO2max
VOC(s)
VP
VP 50%
VPD
VT
VTB
VTmax
VUA
vWF
W
W126
W95
WBC
WBGT
we
WCB
ultraviolet radiation at wavelengths
of 280 to 320 nm
ultraviolet radiation at wavelengths
of 200 to 280 nm
Ultraviolet Differential Absorption
Lidar
vanadium
volt, millivolt, microvolt
alveolar ventilation
valine
vital capacity
vascular cell adhesion molecule
deposition rate, deposition velocity
(cm/sec)
volume of the anatomic or
physiological dead space
ventilation rate; minute ventilation;
ventilatory volume
vascular endothelial growth factor
maximum minute ventilation
maximum velocity
maximum expiratory flow at 25%
of the vital capacity
maximum expiratory flow at 50%
of the vital capacity
maximum expiratory flow at 75%
of the vital capacity
volume median diameter
nasal volume
oxygen consumption
maximum volume per time, of
oxygen (maximal oxygen
consumption, maximal oxygen
uptake or aerobic capacity)
volatile organic compound(s)
volumetric penetration
volume at which 50% of an inhaled
bolus is absorbed
vapor pressure deficit; Vehicles
per day; Ventricular premature
depolarization
tidal volume
terminal bronchiole region volume
maximum tidal volume
volume of the upper airways
von Willebrand factor
width; wilderness; week(s)
cumulative integrated exposure
index with a sigmoidal weighting
function
cumulative integrated exposure
index with a sigmoidal weighting
function
white blood cell
wet bulb globe temperature
sigmoidal weighting of hourly O3
concentration
warm conveyor belt
WED
WF, WFM
WHI
WHO
W/m2, W m'2
WMO
WMO/UNEP
WRF
Ws
WS
WT
wt %
WUS
w/v
Y
yr
Z
ZAPS
ZELIG
Zn
(U.S. EPA NHEERL) Western
Ecology Division
White Face Mountain site
Women's Health Initiative
World Health Organization
watts per square meter
World Meteorological Organization
World Meteorological
Organization/United Nations
Environment Program
Weather Research and
Forecasting model
Wassilewskija Arabidopsis ecotype
wood smoke
wild type; White Top Mountain site
percent by weight
western U.S.
weight per volume
three parameter Weibull model
year
Airway generation
Zonal Air Pollution System
a forest succession simulation
model
-------
PREAMBLE
Process of ISA Development
This preamble outlines the general process for developing an Integrated Science
Assessment (ISA) including the framework for evaluating weight of evidence and
drawing scientific conclusions and causal judgments. The ISA provides a concise
review, synthesis, and evaluation of the most policy-relevant science to serve as a
scientific foundation for the review of the National Ambient Air Quality Standards
(NAAQS). The general process for NAAQS reviews is described at
http://www.epa.gov/ttn/naaqs/review.html. Figure I depicts the general NAAQS review
process and information for individual NAAQS reviews is available at
www.epa.gov/ttn/naaqs. This preamble is a general discussion of the basic steps and
criteria used in developing an ISA; for each ISA, specific details and considerations
are included in the introductory section for that assessment.
The fundamental process for developing an ISA includes:
• literature searches;
• study selection;
• evaluation and integration of the evidence;
• development of scientific conclusions and causal judgments.
An initial step in this process is publication of a call for information in the Federal
Register that invites the public to provide information relevant to the assessment,
such as new or recent publications on health or welfare1 effects of the pollutant, or
from atmospheric and exposure sciences fields. EPA maintains an ongoing literature
search process for identification of relevant scientific studies published since the last
review of the NAAQS. Search strategies are designed for pollutants and scientific
disciplines and iteratively modified to optimize identification of pertinent
publications. Papers are identified for inclusion in several additional ways:
specialized searches on specific topics; independent review of tables of contents for
journals in which relevant papers may be published; independent identification of
relevant literature by expert scientists; review of citations in previous assessments
and identification by the public and the Clean Air Scientific Advisory Committee
(CASAC) during the external review process. This literature search and study
selection process is depicted in Figure II. Publications considered for inclusion in the
ISA are added to the Health and Environmental Research Online (HERO) database
developed by EPA (http://hero.epa.gov/): the references in the ISA include a hyperlink to
the database.
1 Welfare effects as defined in Clean Air Act (CAA) section 302(h) [42 U.S.C. 7602(h)] include, but are not limited to, "effects on
soils, water, crops, vegetation, man-made materials, animals, wildlife, weather, visibility and climate, damage to and deterioration
of property, and hazards to transportation, as well as effects on economic values and on personal comfort and well-being."
li
-------
Studies that have undergone scientific peer review and have been published or
accepted for publication and reports that have undergone review are considered for
inclusion in the ISA. Analyses conducted by EPA using publicly available data are
also considered for inclusion in the ISA. All relevant epidemiologic, controlled
human exposure, toxicological, and ecological and welfare effects studies published
since the last review are considered, including those related to exposure-response
relationships, mode(s) of action (MOA), and potentially at-risk populations and
lifestages. Studies on atmospheric chemistry, environmental fate and transport,
dosimetry, toxicokinetics and exposure are also considered for inclusion in the
document, as well as analyses of air quality and emissions data. References that were
considered for inclusion in a specific ISA can be found using the HERO website
(http://hero.epa.gov).
Integrated Review Plan (IRP) (imeline and key
policy-relevant issues and scientific questions
Integrated Science Assessment (ISA): evaluation and
synthesis of most policy-relevant studies
JL
Risk/Exposure Assessment (REA):
quantitative assessment, as warranted, focused
on key results, observations, and uncertainties
Policy Assessment (PA): staff analysis of
policy options based on integration and
interpretation of information in the ISA and REA
EPA ,
proposed
decisions on
> standards
Agency decision
making and draft
final notice
Interagency
review
Clean Air Scientific
Advisory Committee
(CASAC) review
Public comment
Figure I Illustration of the key steps in the process of the review of National
Ambient Air Quality Standards.
Hi
-------
Literature
Search
Strategies
Citations from
Past Assessments
Peer Review
Recommenda tions
Criteria for study evaluation include:
• Are the study populations, subjects,
or animal models adequately
selected, and are they sufficiently
well defined to allow for meaningful
comparisons between study or
exposure groups?
• Are the statistical analyses
appropriate, properly performed, and
properly interpreted? Are likely
covariates adequately controlled or
taken into account in the study
design and statistical analysis?
• Are the air quality data, exposure, or
dose metrics of adequate quality and
sufficiently representative of
information regarding ambient
conditions?
• Are the health, ecological or welfare
effect measurements meaningful,
valid and reliable?
• Do the analytical methods provide
adequate sensitivity and precision to
support conclusions?
Figure II Illustration of processes for literature search and study selection
used for development of ISAs.
Each ISA builds upon the conclusions of previous assessments for the pollutant
under review. EPA focuses on peer reviewed literature published following the
completion of the previous review (2006 O3 AQCD) and on any new interpretations
of previous literature, integrating the results of recent scientific studies with previous
findings. Important earlier studies may be discussed in detail to reinforce key
concepts and conclusions or for reinterpretation in light of newer data. Earlier studies
also are the primary focus in some areas of the document where research efforts have
subsided, or if these earlier studies remain the definitive works available in the
literature.
Selection of studies for inclusion in the ISA is based on the general scientific quality
of the study, and consideration of the extent to which the study is informative and
policy-relevant. Policy relevant and informative studies include those that provide a
basis for or describe the relationship between the criteria pollutant and effects,
including studies that offer innovation in method or design and studies that reduce
uncertainty on critical issues, such as analyses of confounding or effect modification
by copollutants or other variables, analyses of concentration-response or dose-
liii
-------
response relationships, or analyses related to time between exposure and response.
Emphasis is placed on studies that examine effects associated with pollutant
concentrations relevant to current population and ecosystem exposures, and
particularly those pertaining to concentrations currently found in ambient air. Other
studies are included if they contain unique data, such as a previously unreported
effect or MOA for an observed effect, or examine multiple concentrations to
elucidate exposure-response relationships. In general, in assessing the scientific
quality and relevance of health and welfare effects studies, the following
considerations have been taken into account when selecting studies for inclusion in
the ISA.
• Are the study populations, subjects, or animal models adequately selected, and
are they sufficiently well defined to allow for meaningful comparisons
between study or exposure groups?
• Are the statistical analyses appropriate, properly performed, and properly
interpreted? Are likely covariates adequately controlled or taken into account
in the study design and statistical analysis?
• Are the air quality data, exposure, or dose metrics of adequate quality and
sufficiently representative of information regarding ambient conditions?
• Are the health, ecological or welfare effect measurements meaningful, valid
and reliable?
• Do the analytical methods provide adequate sensitivity and precision to
support conclusions?
Considerations specific to particular disciplines include the following: In selecting
epidemiologic studies, EPA considers whether a given study: (1) presents
information on associations with short- or long-term pollutant exposures at or near
conditions relevant to ambient exposures; (2) addresses potential confounding by
other pollutants; (3) assesses potential effect modifiers; (4) evaluates health
endpoints and populations not previously extensively researched; and (5) evaluates
important methodological issues related to interpretation of the health evidence
(e.g., lag or time period between exposure and effects, model specifications,
thresholds, mortality displacement).
Considerations for the selection of research evaluating controlled human exposure or
animal toxicological studies include a focus on studies conducted using relevant
pollutant exposures. For both types of studies, relevant pollutant exposures are
considered to be those generally within one or two orders of magnitude of ambient
concentrations. Studies in which higher doses were used may also be considered if
they provide information relevant to understanding MOA or mechanisms, as noted
below.
Evaluation of controlled human exposure studies focuses on those that approximated
expected human exposure conditions in terms of concentration and duration. Studies
should include control exposures to filtered air, as appropriate. In the selection of
controlled human exposure studies, emphasis is placed on studies that: (1) investigate
potentially at-risk populations and lifestages such as people with asthma or
liv
-------
cardiovascular diseases, children or older adults; (2) address issues such as
concentration-response or time-course of responses; and (3) have sufficient statistical
power to assess findings.
Review of the animal toxicological evidence focuses on studies that approximate
expected human dose conditions, which vary depending on the dosimetry,
toxicokinetics, and biological sensitivity of the particular laboratory animal species
or strains studied. Emphasis is placed on studies that: (1) investigate animal models
of disease that can provide information on populations potentially at increased risk of
effects; (2) address issues such as concentration-response or time-course of
responses; and (3) have sufficient statistical power to assess findings. Due to
resource constraints on exposure duration and numbers of animals tested, animal
studies typically utilize high-concentration exposures to acquire data relating to
mechanisms and assure a measurable response. Emphasis is placed on studies using
doses or concentrations generally within 1-2 orders of magnitude of current levels.
Studies with higher concentration exposures or doses are considered to the extent that
they provide useful information to inform understanding of interspecies differences
between healthy and at-risk human populations. Results from in vitro studies may
also be included if they provide mechanistic insight or further support for results
demonstrated in vivo.
These criteria provide benchmarks for evaluating various studies and for focusing on
the policy-relevant studies in assessing the body of health, ecological and welfare
effects evidence. As stated initially, the intent of the ISA is to provide a concise
review, synthesis, and evaluation of the most policy-relevant science to serve as a
scientific foundation for the review of the NAAQS, not extensive summaries of all
health, ecological and welfare effects studies for a pollutant. Of most relevance for
inclusion of studies is whether they provide useful qualitative or quantitative
information on exposure-effect or exposure-response relationships for effects
associated with pollutant exposures at doses or concentrations relevant to ambient
conditions that can inform decisions on whether to retain or revise the standards.
The general process for ISA development is illustrated in Figure III. In developing an
ISA, EPA reviews and summarizes the evidence from studies of atmospheric
sciences; human exposure, toxicological, controlled human exposure and
epidemiologic studies; and studies of ecological and welfare effects. In the process of
developing the first draft ISA, EPA may convene a peer input meeting in which EPA
the scientific content of preliminary draft materials is reviewed to ensure that the ISA
is up to date and focused on the most policy-relevant findings, and to assist EPA with
integration of evidence within and across disciplines. EPA integrates the evidence
from across scientific disciplines or study types and characterizes the weight of
evidence for relationships between the pollutant and various outcomes.
The integration of evidence on health, and ecological or welfare effects, involves
collaboration between scientists from various disciplines. As an example, an
evaluation of health effects evidence would include the integration of the results from
epidemiologic, controlled human exposure, and toxicological studies, and application
of the causal framework (described below) to draw conclusions. Integration of results
Iv
-------
on health or ecological effects that are logically or mechanistically connected (e.g., a
spectrum of effects on the respiratory system) informs judgments of causality. Using
the causal framework described in the following section, EPA scientists consider
aspects such as strength, consistency, coherence, and biological plausibility of the
evidence, and develop causality determinations on the nature of the relationships.
Causality determinations often entail an iterative process of review and evaluation of
the evidence. Two drafts of the ISA are typically released for review by the CAS AC
and the public, and comments received on the characterization of the science as well
as the implementation of the causal framework are carefully considered in revising
and completing the final ISA.
Ivi
-------
Integrated Science Assessment Development Process
Literature Search and
Study Selection
(See Figure II)
Characterization of Evidence
Develop initial sections or "building blocks" of scientific evidence for assessment:
review and summarize new study results, by outcome/effect category and
discipline, for example, toxicological studies of lung function. Summarize findings
and conclusions from previous assessment. As appropriate, develop initial
conclusions about the available evidence.
Peer Input Consultation
Review of initial draft materials for scientific quality of "building blocks" from
scientists from both outside and within EPA; also includes discussion of results to
facilitate integration of findings. Structure varies; may be public meeting or public
teleconference organized by contractor.
Evaluation, Synthesis, and Integration of Evidence
Integrate evidence from scientific disciplines orstudy types - for example,
toxicological, controlled human exposure and epidemiologic study findings for
particular health outcome. Evaluate evidence for related groups of endpoints or
outcomes to draw conclusions regarding health or welfare effect categories.
Development of Conclusions and Causal Determinations
Evaluate weight of evidence and develop judgments regarding causality for health
or welfare effect categories, integrating health or welfare effects evidence with
information on mode of action and exposure assessment. Develop conclusions
regarding concentration- or dose-response relationships, potentially at-risk
populations or ecosystems.
Draft Integrated Science Assessment
Final Integrated Science Assessment
Clean Air Scientific Advisory Committee
Review in public meeting; anticipated review of two
drafts of ISA
Public Comments
Comments on draft ISA solicited by EPA
Figure
Characterization of the general process of ISA development.
Ivii
-------
EPA Framework for Causal Determination
EPA has developed a consistent and transparent basis for integration of scientific
evidence and evaluation of the causal nature of air pollution-related health or welfare
effects for use in developing ISAs. The framework described below establishes
uniform language concerning causality and brings more specificity to the findings.
This standardized language was drawn from sources across the federal government
and wider scientific community, especially the National Academy of Sciences (NAS)
Institute of Medicine (IOM) document, Improving the Presumptive Disability
Decision-Making Process for Veterans (Samet and Bodurow. 2008). a
comprehensive report on evaluating causality. This framework:
• describes the kinds of scientific evidence used in establishing a general causal
relationship between exposure and health effects;
• characterizes the process for integration and evaluation of evidence necessary
to reach a conclusion about the existence of a causal relationship;
• identifies issues and approaches related to uncertainty;
• provides a framework for classifying and characterizing the weight of
evidence in support of a general causal relationship.
Approaches to assessing the separate and combined lines of evidence
(e.g., epidemiologic, controlled human exposure, and animal toxicological studies)
have been formulated by a number of regulatory and science agencies, including the
IOM of the NAS (Samet and Bodurow. 2008). International Agency for Research on
Cancer (IARC. 2006). U.S. EPA (2005). and Centers for Disease Control and
Prevention (CDC. 2004). Causal inference criteria have also been described for
ecological effects evidence (U.S. EPA. 1998: Fox. 1991). These formalized
approaches offer guidance for assessing causality. The frameworks are similar in
nature, although adapted to different purposes, and have proven effective in
providing a uniform structure and language for causal determinations.
Evaluating Evidence for Inferring Causation
The 1964 Surgeon General's (U.S. Department of Health, Education and Welfare
[HEW]) report on tobacco smoking defined "cause" as a "significant, effectual
relationship between an agent and an associated disorder or disease in the host"
(HEW. 1964). More generally, a cause is defined as an agent that brings about an
effect or a result. An association is the statistical relationship among variables; alone,
however, it is insufficient proof of a causal relationship between an exposure and a
health outcome. Unlike an association, a causal claim supports the creation of
counterfactual claims, that is, a claim about what the world would have been like
under different or changed circumstances (Samet and Bodurow. 2008).
Many of the health and environmental outcomes reported in these studies have
complex etiologies. Diseases such as asthma, coronary heart disease (CHD) or cancer
Iviii
-------
are typically initiated by multiple agents. Outcomes depend on a variety of factors,
such as age, genetic susceptibility, nutritional status, immune competence, and social
factors (Samet and Bodurow, 2008; Gee and Payne-Sturges, 2004). Effects on
ecosystems are often also multifactorial with a complex web of causation. Further,
exposure to a combination of agents could cause synergistic or antagonistic effects.
Thus, the observed risk may represent the net effect of many actions and
counteractions.
Scientific findings incorporate uncertainty. "Uncertainty" can be defined as having
limited knowledge to exactly describe an existing state or future outcome, e.g., the
lack of knowledge about the correct value for a specific measure or estimate.
Uncertainty analysis may be qualitative or quantitative in nature. In many cases, the
analysis is qualitative, and can include professional judgment or inferences based on
analogy with similar situations. Quantitative uncertainty analysis may include use of
simple measures (e.g., ranges) and analytical techniques. Quantitative uncertainty
analysis might progress to more complex measures and techniques, if needed for
decision support. Various approaches to evaluating uncertainty include classical
statistical methods, sensitivity analysis, or probabilistic uncertainty analysis, in order
of increasing complexity and data requirements. However, data may not be available
for all aspects of an assessment and those data that are available may be of
questionable or unknown quality. Ultimately, the assessment is based on a number of
assumptions with varying degrees of uncertainty.
Publication bias is a source of uncertainty regarding the magnitude of health risk
estimates. It is well understood that studies reporting non-null findings are more
likely to be published than reports of null findings. Publication bias can result in
overestimation of effect estimate sizes (loannidis. 2008). For example, effect
estimates from single-city epidemiologic studies have been found to be generally
larger than those from multicity studies which is an indication of publication bias in
that null or negative single-city results may be reported in a multicity analyses but
might not be published independently (Bell et al., 2005).
Consideration of Evidence from Scientific Disciplines
Moving from association to causation involves the elimination of alternative
explanations for the association. The ISA focuses on evaluation of the findings from
the body of evidence, drawing upon the results of all studies determined to meet the
criteria described previously. Causality determinations are based on the evaluation,
integration, and synthesis of evidence from across scientific disciplines. The relative
importance of different types of evidence varies by pollutant or assessment, as does
the availability of different types of evidence for causality determination. Three
general types of studies inform consideration of human health effects: controlled
human exposure, epidemiologic, and toxicological studies. Evidence on ecological or
welfare effects may be drawn from a variety of experimental approaches (e.g.,
lix
-------
greenhouse, laboratory, field) and numerous disciplines (e.g., community ecology,
biogeochemistry, and paleontological/historical reconstructions).
Direct evidence of a relationship between pollutant exposures and human health
effects comes from controlled human exposure studies. Such studies experimentally
evaluate the health effects of administered exposures in human volunteers under
highly controlled laboratory conditions. Also referred to as human clinical studies,
these experiments allow investigators to expose subjects to known concentrations of
air pollutants under carefully regulated environmental conditions and activity levels.
These studies provide important information on the biological plausibility of
associations observed in epidemiologic studies. Essential dose-response profiles and
ranges of response severity can be established with these studies. In some instances,
controlled human exposure studies can also be used to characterize
concentration-response relationships at pollutant concentrations relevant to ambient
conditions. Controlled human exposures are typically conducted using a randomized
crossover design, with subjects exposed both to the pollutant and a clean air control.
In this way, subjects serve as their own controls, effectively controlling for many
potential confounders. Considerations for evaluating controlled human study findings
include the generally small sample size and short exposure time used in experimental
studies, and that severe health outcomes are not assessed. By experimental design,
controlled human exposure studies are structured to evaluate physiological or
biomolecular outcomes in response to exposure to a specific air pollutant and/or
combination of pollutants. In addition, the study design generally precludes inclusion
of subjects with serious health conditions, and therefore the results often cannot be
generalized to an entire population. Although some controlled human exposure
studies have included health-compromised individuals such as those with respiratory
or cardiovascular disease, these individuals may also be relatively healthy and may
not represent the most sensitive individuals in the population. Thus, observed effects
in these studies may underestimate the response in certain populations.
Epidemiologic studies provide important information on the associations between
health effects and exposure of human populations to ambient air pollution.
In epidemiologic or observational studies of humans, the investigator generally does
not control exposures or intervene with the study population. Broadly, observational
studies can describe associations between exposures and effects. These studies fall
into several categories: e.g., cross-sectional, prospective cohort, panel, and
time-series studies. Cross-sectional studies use health outcome, exposure and
covariate data available at the community level (e.g., annual mortality rates and
pollutant concentrations), but do not have individual-level data. Prospective cohort
studies have some data collected at the individual level, generally health outcome
data, and in some cases individual-level data on exposure and covariates are
collected. Time-series studies evaluate the relationship for changes in a health
outcome with changes in exposure indicators, such as an association between daily
changes in mortality with air pollution. Panel studies include repeated measurements
of health outcomes, such as respiratory symptoms or heart rhythm variable, at the
individual level. "Natural experiments" offer the opportunity to investigate changes
in health related to a change in exposure, such as closure of a pollution source.
Ix
-------
In evaluating epidemiologic studies, consideration of many study design factors and
issues must be taken into account to properly inform their interpretation. One key
consideration is evaluation of the potential contribution of the pollutant to a health
outcome when it is a component of a complex air pollutant mixture. Reported effect
estimates in epidemiologic studies may reflect: independent effects on health
outcomes; effects of the pollutant acting as an indicator of a copollutant or a complex
ambient air pollution mixture; effects resulting from interactions between that
pollutant and copollutants.
In the evaluation of epidemiologic evidence, one important consideration is potential
confounding. Confounding is "... a confusion of effects. Specifically, the apparent
effect of the exposure of interest is distorted because the effect of an extraneous
factor is mistaken for or mixed with the actual exposure effect (which may be null)"
(Rothman and Greenland, 1998). One approach to remove spurious associations (due
to possible confounders); is to control for characteristics that may differ between
exposed and unexposed persons; this is frequently termed "adjustment." Scientific
judgment is needed to evaluate likely sources and extent of confounding, together
with consideration of how well the existing constellation of study designs, results,
and analyses address the potential for erroneous inferences. A confounder is
associated with both the exposure and the effect; for example, confounding can occur
between correlated pollutants that are associated with the same effect.
Several statistical methods are available to detect and control for potential
confounders, with none of them being completely satisfactory. Multivariable
regression models constitute one tool for estimating the association between
exposure and outcome after adjusting for characteristics of participants that might
confound the results. The use of multipollutant regression models has been the
prevailing approach for controlling potential confounding by copollutants in air
pollution health effects studies. Finding the likely causal pollutant from
multipollutant regression models is made difficult by the possibility that one or more
air pollutants may be acting as a surrogate for an unmeasured or poorly measured
pollutant or for a particular mixture of pollutants. In addition, pollutants may
independently exert effects on the same system; for example, several pollutants may
be associated with respiratory effects through either the same or different modes of
action. The number and degree of diversity of covariates, as well as their relevance to
the potential confounders, remain matters of scientific judgment. Despite these
limitations, the use of multipollutant models is still the prevailing approach employed
in most air pollution epidemiologic studies and provides some insight into the
potential for confounding or interaction among pollutants.
Confidence that unmeasured confounders are not producing the findings is increased
when multiple studies are conducted in various settings using different subjects or
exposures, each of which might eliminate another source of confounding from
consideration. For example, multicity studies can provide insight on potential
confounding through the use of a consistent method to analyze data from across
locations with different levels of copollutants and other covariates. Intervention
Ixi
-------
studies, because of their quasi-experimental nature, can be particularly useful in
characterizing causation.
Another important consideration in the evaluation of epidemiologic evidence is effect
modification, which occurs when the effect differs between subgroups or strata; for
example, effect estimates that vary by age group or potential risk factor. As stated by
Rothman and Greenland (1998):
"Effect-measure modification differs from confounding in several ways.
The main difference is that, whereas confounding is a bias that the
investigator hopes to prevent or remove from the effect estimate, effect-
measure modification is a property of the effect under study ...
In epidemiologic analysis one tries to eliminate confounding but one tries to
detect and estimate effect-measure modification."
When a risk factor is a confounder, it is the true cause of the association observed
between the exposure and the outcome; when a risk factor is an effect modifier, it
changes the magnitude of the association between the exposure and the outcome in
stratified analyses. For example, the presence of a pre-existing disease or indicator of
low socioeconomic status may act as effect modifiers if they are associated with
increased risk of effects related to air pollution exposure. It is often possible to
stratify the relationship between health outcome and exposure by one or more of
these potential effect modifiers. For variables that modify the association, effect
estimates in each stratum will be different from one another and different from the
overall estimate, indicating a different exposure-response relationship may exist in
populations represented by these variables.
Exposure measurement error, which refers to the uncertainty associated with the
exposure metrics used to represent exposure of an individual or population, can be an
important contributor to uncertainty in air pollution epidemiologic study results.
Exposure error can influence observed epidemiologic associations between ambient
pollutant concentrations and health outcomes by biasing effect estimates toward or
away from the null and widening confidence intervals around those estimates (Zeger
et al., 2000). There are several components that contribute to exposure measurement
error in air pollution epidemiologic studies, including the difference between true and
measured ambient concentrations, the difference between average personal exposure
to ambient pollutants and ambient concentrations at central monitoring sites, and the
use of average population exposure rather than individual exposure estimates.
Factors that could influence exposure estimates include nonambient sources of
exposure, topography of the natural and built environment, meteorology,
measurement errors, time-location-activity patterns, and the extent to which ambient
pollutants penetrate indoor environments. The importance of exposure error varies
with study design and is dependent on the spatial and temporal aspects of the design.
The third main type of health effects evidence, animal toxicological studies, provides
information on the pollutant's biological action under controlled and monitored
exposure circumstances. Taking into account physiological differences of the
experimental species from humans, these studies inform characterization of health
Ixii
-------
effects of concern, exposure-response relationships and MO As. Further, animal
models can inform determinations of at-risk populations. These studies evaluate the
effects of exposures to a variety of pollutants in a highly controlled laboratory setting
and allow exploration of toxicological pathways or mechanisms by which a pollutant
may cause effects. Understanding the biological mechanisms underlying various
health outcomes can prove crucial in establishing or negating causality. In the
absence of human studies data, extensive, well-conducted animal toxicological
studies can support determinations of causality, if the evidence base indicates that
similar responses are expected in humans under ambient exposure conditions.
Interpretations of animal toxicological studies are affected by limitations associated
with extrapolation between animal and human responses. The differences between
humans and other species have to be taken into consideration, including metabolism,
hormonal regulation, breathing pattern, and differences in lung structure and
anatomy. Also, in spite of a high degree of homology and the existence of a high
percentage of orthologous genes across humans and rodents (particularly mice),
extrapolation of molecular alterations at the gene level is complicated by species-
specific differences in transcriptional regulation. Given these differences, there are
uncertainties associated with quantitative extrapolations of observed
pollutant-induced pathophysiological alterations between laboratory animals and
humans, as those alterations are under the control of widely varying biochemical,
endocrine, and neuronal factors.
For ecological effects assessment, both laboratory and field studies (including field
experiments and observational studies) can provide useful data for causality
determination. Because conditions can be controlled in laboratory studies, responses
may be less variable and smaller differences may be easier to detect. However, the
control conditions may limit the range of responses (e.g., animals may not be able to
seek alternative food sources) or incompletely reflect pollutant bioavailability, so
they may not reflect responses that would occur in the natural environment.
In addition, larger-scale processes are difficult to reproduce in the laboratory.
Field observational studies measure biological changes in uncontrolled situations,
and describe an association between a disturbance and an ecological effect. Field data
can provide important information for assessments of multiple stressors or where
site-specific factors significantly influence exposure. They are also often useful for
analyses of larger geographic scales and higher levels of biological organization.
However, because conditions are not controlled, variability is expected to be higher
and differences harder to detect. Field surveys are most useful for linking stressors
with effects when stressor and effect levels are measured concurrently. The presence
of confounding factors can make it difficult to attribute observed effects to specific
stressors.
Intermediate between laboratory and field are studies that use environmental media
collected from the field to examine response in the laboratory, and experiments that
are performed in the natural environment while controlling for some environmental
conditions (i.e., mesocosm studies). This type of study in manipulated natural
environments can be considered a hybrid between a field experiment and laboratory
Ixiii
-------
study since some aspects are performed under controlled conditions but others are
not. They make it possible to observe community and/or ecosystem dynamics, and
provide strong evidence for causality when combined with findings of studies that
have been made under more controlled conditions.
Application of Framework for Causal Determination
In its evaluation and integration of the scientific evidence on health or welfare effects
of criteria pollutants, EPA determines the weight of evidence in support of causation
and characterizes the strength of any resulting causal classification. EPA also
evaluates the quantitative evidence and draws scientific conclusions, to the extent
possible, regarding the concentration-response relationships and the loads to
ecosystems, exposures, doses or concentrations, exposure duration, and pattern of
exposures at which effects are observed.
To aid judgment, various "aspects"1 of causality have been discussed by many
philosophers and scientists. The 1964 Surgeon General's report on tobacco smoking
discussed criteria for the evaluation of epidemiologic studies, focusing on
consistency, strength, specificity, temporal relationship, and coherence (HEW, 1964).
Sir Austin Bradford Hill (Hill. 1965) articulated aspects of causality in epidemiology
and public health that have been widely used (Samet and Bodurow. 2008: IARC.
2006: U.S. EPA. 2005: CDC. 2004). These aspects (Hill. 1965) have been modified
(Table I) for use in causal determinations specific to health and welfare effects for
pollutant exposures (U.S. EPA. 2009d).2 Although these aspects provide a
framework for assessing the evidence, they do not lend themselves to being
considered in terms of simple formulas or fixed rules of evidence leading to
conclusions about causality (Hill. 1965). For example, one cannot simply count the
number of studies reporting statistically significant results or statistically
nonsignificant results and reach credible conclusions about the relative weight of the
evidence and the likelihood of causality. Rather, these aspects provide a framework
for systematic appraisal of the body of evidence, informed by peer and public
comment and advice, which includes weighing alternative views on controversial
issues. In addition, it is important to note that the aspects in Table I cannot be used as
a strict checklist, but rather to determine the weight of the evidence for inferring
causality. In particular, not meeting one or more of the principles does not
automatically preclude a determination of causality [see discussion in CDC (2004)].
1 The "aspects" described by Sir Austin Bradford Hill (Hill, 1965) have become, in the subsequent literature, more commonly
described as "criteria." The original term "aspects" is used here to avoid confusion with "criteria" as it is used, with different
meaning, in the Clean Air Act.
2 The Hill aspects were developed for interpretation of epidemiologic results. They have been modified here for use with a broader
array of data, i.e., epidemiologic, controlled human exposure, ecological, and animal toxicological studies, as well as in vitro data,
and to be more consistent with the U.S. EPA (2005) Guidelines for Carcinogen Risk Assessment.
Ixiv
-------
Table I
Aspects to aid in judging causality.
Aspect
Description
Consistency of the
observed association
An inference of causality is strengthened when a pattern of elevated risks is observed
across several independent studies. The reproducibility of findings constitutes one of the
strongest arguments for causality. If there are discordant results among investigations,
possible reasons such as differences in exposure, confounding factors, and the power
of the study are considered.
Coherence An inference of causality from one line of evidence (e.g., epidemiologic, controlled
human exposure [clinical], or animal studies) may be strengthened by other lines of
evidence that support a cause-and-effect interpretation of the association. Evidence on
ecological or welfare effects may be drawn from a variety of experimental approaches
(e.g., greenhouse, laboratory, and field) and subdisciplines of ecology (e.g., community
ecology, biogeochemistry, and paleontological/historical reconstructions).
The coherence of evidence from various fields greatly adds to the strength of an
inference of causality. In addition, there may be coherence in demonstrating effects
across multiple study designs or related health endpoints within one scientific line of
evidence.
Biological plausibility
Biological gradient
(exposure-response
relationship)
Strength of the observed
association
Experimental evidence
Temporal relationship of
the observed association
An inference of causality tends to be strengthened by consistency with data from
experimental studies or other sources demonstrating plausible biological mechanisms.
A proposed mechanistic linking between an effect and exposure to the agent is an
important source of support for causality, especially when data establishing the
existence and functioning of those mechanistic links are available.
A well-characterized exposure-response relationship (e.g., increasing effects associated
with greater exposure) strongly suggests cause and effect, especially when such
relationships are also observed for duration of exposure (e.g., increasing effects
observed following longer exposure times).
The finding of large, precise risks increases confidence that the association is not likely
due to chance, bias, or other factors. However, it is noted that a small magnitude in an
effect estimate may represent a substantial effect in a population.
Strong evidence for causality can be provided through "natural experiments" when a
change in exposure is found to result in a change in occurrence or frequency of health
or welfare effects.
Evidence of a temporal sequence between the introduction of an agent, and appearance
of the effect, constitutes another argument in favor of causality.
Specificity of the
observed association
Analogy
Evidence linking a specific outcome to an exposure can provide a strong argument for
causation. However, it must be recognized that rarely, if ever, does exposure to a
pollutant invariably predict the occurrence of an outcome, and that a given outcome may
have multiple causes.
Structure activity relationships and information on the agent's structural analogs can
provide insight into whether an association is causal. Similarly, information on mode of
action for a chemical, as one of many structural analogs, can inform decisions regarding
likely causality.
Ixv
-------
Determination of Causality
In the ISA, EPA assesses the body of relevant literature, building upon evidence
available during previous NAAQS reviews, to draw conclusions on the causal
relationships between relevant pollutant exposures and health or environmental
effects. ISAs use a five-level hierarchy that classifies the weight of evidence for
causation1. In developing this hierarchy, EPA has drawn on the work of previous
evaluations, most prominently the lOM's Improving the Presumptive Disability
Decision-Making Process for Veterans (Samet and Bodurow. 2008). EPA's
Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005). and the U.S. Surgeon
General's smoking report (CDC. 2004). This weight of evidence evaluation is based
on integration of findings from various lines of evidence from across the health and
environmental effects disciplines. These separate judgments are integrated into a
qualitative statement about the overall weight of the evidence and causality. The five
descriptors for causal determination are described in Table II.
Determination of causality involves the evaluation and integration of evidence for
different types of health, ecological or welfare effects associated with short- and
long-term exposure periods. In making determinations of causality, evidence is
evaluated for major outcome categories or groups of related endpoints
(e.g., respiratory effects, vegetation growth), integrating evidence from across
disciplines, and assessing the coherence of evidence across a spectrum of related
endpoints to draw conclusions regarding causality. In discussing the causal
determination, EPA characterizes the evidence on which the judgment is based,
including strength of evidence for individual endpoints within the outcome category
or group of related endpoints.
In drawing judgments regarding causality for the criteria air pollutants, the ISA
focuses on evidence of effects in the range of relevant pollutant exposures or doses,
and not on determination of causality at any dose. Emphasis is placed on evidence of
effects at doses (e.g., blood Pb concentration) or exposures (e.g., air concentrations)
that are relevant to, or somewhat above, those currently experienced by the
population. The extent to which studies of higher concentrations are considered
varies by pollutant and major outcome category, but generally includes those with
doses or exposures in the range of one to two orders of magnitude above current or
ambient conditions. Studies that use higher doses or exposures may also be
considered to the extent that they provide useful information to inform understanding
of mode of action, interspecies differences, or factors that may increase risk of effects
for a population. Thus, a causality determination is based on weight of evidence
evaluation for health, ecological or welfare effects, focusing on the evidence from
exposures or doses generally ranging from current levels to one or two orders of
magnitude above current levels.
1 Both the CDC and IOM frameworks use a four-category hierarchy for the strength of the evidence. A five-level hierarchy is used
here to be consistent with the EPA Guidelines for Carcinogen Risk Assessment and to provide a more nuanced set of categories.
Ixvi
-------
In addition, EPA evaluates evidence relevant to understand the quantitative
relationships between pollutant exposures and health, ecological or welfare effects.
This includes evaluation of the form of concentration-response or dose-response
relationships and, to the extent possible, drawing conclusions on the levels at which
effects are observed. The ISA also draws scientific conclusions regarding important
exposure conditions for effects and populations that may be at greater risk for effects,
as described in the following section.
Ixvii
-------
Table II
Weight of evidence for causal determination.
Health Effects
Ecological and Welfare Effects
Causal Evidence is sufficient to conclude that there is a causal
relationship relationship with relevant pollutant exposures
(i.e., doses or exposures generally within one to two
orders of magnitude of current levels). That is, the
pollutant has been shown to result in health effects in
studies in which chance, bias, and confounding could
be ruled out with reasonable confidence. For
example: a) controlled human exposure studies that
demonstrate consistent effects; or b) observational
studies that cannot be explained by plausible
alternatives or are supported by other lines of
evidence (e.g., animal studies or mode of action
information). Evidence includes multiple high-quality
studies
Evidence is sufficient to conclude that there is a causal
relationship with relevant pollutant exposures
i.e., doses or exposures generally within one to two
orders of magnitude of current levels). That is, the
pollutant has been shown to result in effects in studies
in which chance, bias, and confounding could be ruled
out with reasonable confidence. Controlled exposure
studies (laboratory or small- to medium-scale field
studies) provide the strongest evidence for causality,
but the scope of inference may be limited. Generally,
determination is based on multiple studies conducted
by multiple research groups, and evidence that is
considered sufficient to infer a causal relationship is
usually obtained from the joint consideration of many
lines of evidence that reinforce each other.
Likely to be a
causal
relationship
Evidence is sufficient to conclude that a causal
relationship is likely to exist with relevant pollutant
exposures, but important uncertainties remain. That
is, the pollutant has been shown to result in health
effects in studies in which chance and bias can be
ruled out with reasonable confidence but potential
issues remain. For example: a) observational studies
show an association, but copollutant exposures are
difficult to address and/or other lines of evidence
(controlled human exposure, animal, or mode of
action information) are limited or inconsistent; or b)
animal toxicological evidence from multiple studies
from different laboratories that demonstrate effects,
but limited or no human data are available. Evidence
generally includes multiple high-quality studies.
Evidence is sufficient to conclude that there is a likely
causal association with relevant pollutant exposures.
That is, an association has been observed between
the pollutant and the outcome in studies in which
chance, bias, and confounding are minimized, but
uncertainties remain. For example, field studies show
a relationship, but suspected interacting factors
cannot be controlled, and other lines of evidence are
limited or inconsistent. Generally, determination is
based on multiple studies in multiple research groups.
Suggestive of
a causal
relationship
Evidence is suggestive of a causal relationship with
relevant pollutant exposures, but is limited. For
example, (a) at least one high-quality epidemiologic
study shows an association with a given health
outcome but the results of other studies are
inconsistent; or (b) a well-conducted toxicological
study, such as those conducted in the National
Toxicology Program (NTP), shows effects in animal
species,
Evidence is suggestive of a causal relationship with
relevant pollutant exposures, but chance, bias and
confounding cannot be ruled out. For example, at
least one high-quality study shows an effect, but the
results of other studies are inconsistent.
Inadequate to
infer a causal
relationship
Evidence is inadequate to determine that a causal
relationship exists with relevant pollutant exposures.
The available studies are of insufficient quantity,
quality, consistency, or statistical power to permit a
conclusion regarding the presence or absence of an
effect.
The available studies are of insufficient quality,
consistency, or statistical power to permit a conclusion
regarding the presence or absence of an effect.
Not likely to
be a causal
relationship
Evidence is suggestive of no causal relationship with
relevant pollutant exposures. Several adequate
studies, covering the full range of levels of exposure
that human beings are known to encounter and
considering at-risk populations, are mutually
consistent in not showing an effect at any level of
exposure.
Several adequate studies, examining relationships with
relevant exposures, are consistent in failing to show
an effect at any level of exposure.
Ixviii
-------
Quantitative Relationships: Effects on Human Populations
Once a determination is made regarding the causal relationship between the pollutant
and outcome category, important questions regarding quantitative relationships
include:
• What is the concentration-response, exposure-response, or dose-response
relationship in the human population?
• What is the interrelationship between incidence and severity of effect?
• What exposure conditions (dose or exposure, duration and pattern) are
important?
• What populations and lifestages appear to be differentially affected (i.e., more
at risk of experiencing effects)?
To address these questions, the entirety of quantitative evidence is evaluated to
characterize pollutant concentrations and exposure durations at which effects were
observed for exposed populations, including populations and lifestages potentially at
increased risk. To accomplish this, evidence is considered from multiple and diverse
types of studies, and a study or set of studies that best approximates the
concentration-response relationships between health outcomes and the pollutant may
be identified. Controlled human exposure studies provide the most direct and
quantifiable exposure-response data on the human health effects of pollutant
exposures. To the extent available, the ISA evaluates results from across
epidemiologic studies that characterize the form of relationships between the
pollutant and health outcomes and draws conclusions on the shape of these
relationships. Animal data may also inform evaluation of concentration-response
relationships, particularly relative to MOAs and characteristics of at-risk populations.
An important consideration in characterizing the public health impacts associated
with exposure to a pollutant is whether the concentration-response relationship is
linear across the range of concentrations or if nonlinear relationships exist along any
part of this range. Of particular interest is the shape of the concentration-response
curve at and below the level of the current standards. Various sources of variability
and uncertainty, such as low data density in the lower concentration range, possible
influence of exposure measurement error, and variability between individuals in
susceptibility to air pollution health effects, tend to smooth and "linearize" the
concentration-response function, and thus can obscure the existence of a threshold or
nonlinear relationship [2006 O3 AQCD (U.S. EPA. 2006b)1. Since individual
thresholds vary from person to person due to individual differences such as genetic
level susceptibility or pre-existing disease conditions (and even can vary from one
time to another for a given person), it can be difficult to demonstrate that a threshold
exists in a population study. These sources of variability and uncertainty may explain
why the available human data at ambient concentrations for some environmental
pollutants (e.g., particulate matter [PM], O3, lead [Pb], environmental tobacco smoke
[ETS], radiation) do not exhibit thresholds for cancer or noncancer health effects,
even though likely mechanisms include nonlinear processes for some key events.
Ixix
-------
Finally, identification of the population groups or lifestages that may be at greater
risk of health effects from air pollutant exposures contributes to an understanding of
the public health impact of pollutant exposures. In the ISA, the term "at-risk
population" is used to encompass populations or lifestages that have a greater
likelihood of experiencing health effects related to exposure to an air pollutant due to
a variety of factors; other terms used in the literature include susceptible, vulnerable,
and sensitive. These factors may be intrinsic, such as genetic or developmental
factors, race, sex, lifestage, or the presence of pre-existing diseases, or they may be
extrinsic, such as socioeconomic status (SES), activity pattern and exercise level,
reduced access to health care, low educational attainment, or increased pollutant
exposures (e.g., near roadways). Epidemiologic studies can help identify populations
potentially at increased risk of effects by evaluating health responses in the study
population. Examples include testing for interactions or effect modification by
factors such as sex, age group, or health status. Experimental studies using animal
models of susceptibility or disease can also inform the extent to which health risks
are likely greater in specific population groups.
Quantitative Relationships: Effects on Ecosystems or Public
Welfare
Key questions for understanding the quantitative relationships between exposure (or
concentration or deposition) to a pollutant and risk to ecosystems or the public
welfare include:
• What elements of the ecosystem (e.g., types, regions, taxonomic groups,
populations, functions, etc.) appear to be affected, or are more sensitive to
effects? Are there differences between locations or materials in welfare effects
responses, such as impaired visibility or materials damage?
• Under what exposure conditions (amount deposited or concentration, duration
and pattern) are effects seen?
• What is the shape of the concentration-response or exposure-response
relationship?
Evaluations of causality generally consider the probability of quantitative changes in
ecological and welfare effects in response to exposure. A challenge to the
quantification of exposure-response relationships for ecological effects is the great
regional and local spatial variability, as well as temporal variability, in ecosystems.
Thus, exposure-response relationships are often determined for a specific ecological
system and scale, rather than at the national or even regional scale. Quantitative
relationships therefore are estimated site by site and may differ greatly between
ecosystems.
Concepts in Evaluating Adversity of Health Effects
In evaluating health evidence, a number of factors can be considered in delineating
between adverse and nonadverse health effects resulting from exposure to air
Ixx
-------
pollution. Some health outcomes, such as hospitalization for respiratory or
cardiovascular diseases, are clearly considered adverse. It is more difficult to
determine the extent of change that constitutes adversity in more subtle health
measures. These include a wide variety of responses, such as alterations in markers
of inflammation or oxidative stress, changes in pulmonary function or heart rate
variability, or alterations in neurocognitive function measures. The challenge is
determining the magnitude of change in these measures when there is no clear point
at which a change becomes adverse. The extent to which a change in health measure
constitutes an adverse health effect may vary between populations. Some changes
that may not be considered adverse in healthy individuals would be potentially
adverse in more at-risk individuals.
The extent to which changes in lung function are adverse has been discussed by the
American Thoracic Society (ATS) in an official statement titled What Constitutes an
Adverse Health Effect of Air Pollution? (ATS. 2000b). An air pollution-induced shift
in the population distribution of a given risk factor for a health outcome was viewed
as adverse, even though it may not increase the risk of any one individual to an
unacceptable level. For example, a population of asthmatics could have a distribution
of lung function such that no identifiable individual has a level associated with
significant impairment. Exposure to air pollution could shift the distribution such that
no identifiable individual experiences any clinically relevant effects. This shift
toward decreased lung function, however, would be considered adverse because
individuals within the population would have diminished reserve function and
therefore would be at increased risk to further environmental insult. The committee
also observed that elevations of biomarkers, such as cell number and types, cytokines
and reactive oxygen species, may signal risk for ongoing injury and clinical effects or
may simply indicate transient responses that can provide insights into mechanisms of
injury, thus illustrating the lack of clear boundaries that separate adverse from
nonadverse effects.
The more subtle health outcomes may be connected mechanistically to health events
that are clearly adverse. For example, air pollution may affect markers of transient
myocardial ischemia such as ST-segment abnormalities and onset of exertional
angina. These effects may not be apparent to the individual, yet may still increase the
risk of a number of cardiac events, including myocardial infarction and sudden death.
Thus, small changes in physiological measures may not appear to be clearly adverse
when considered alone, but may be a part of a coherent and biologically plausible
chain of related health outcomes that range up to responses that are very clearly
adverse, such as hospitalization or mortality.
Concepts in Evaluating Adversity of Ecological Effects
Adversity of ecological effects can be understood in terms ranging in biological level
of organization; from the cellular level to the individual organism and to the
population, community, and ecosystem levels. In the context of ecology, a population
is a group of individuals of the same species, and a community is an assemblage of
populations of different species interacting with one another that inhabit an area.
Ixxi
-------
An ecosystem is the interactive system formed from all living organisms and their
abiotic (physical and chemical) environment within a given area (IPCC, 2007a).
The boundaries of what could be called an ecosystem are somewhat arbitrary,
depending on the focus of interest or study. Thus, the extent of an ecosystem may
range from very small spatial scales to, ultimately, the entire Earth (IPCC, 2007a).
Effects on an individual organism are generally not considered to be adverse to
public welfare. However if effects occur to enough individuals within a population,
then communities and ecosystems may be disrupted. Changes to populations,
communities, and ecosystems can in turn result in an alteration of ecosystem
processes. Ecosystem processes are defined as the metabolic functions of ecosystems
including energy flow, elemental cycling, and the production, consumption and
decomposition of organic matter (U.S. EPA, 2002). Growth, reproduction, and
mortality are species-level endpoints that can be clearly linked to community and
ecosystem effects and are considered to be adverse when negatively affected. Other
endpoints such as changes in behavior and physiological stress can decrease
ecological fitness of an organism, but are harder to link unequivocally to effects at
the population, community, and ecosystem level. The degree to which pollutant
exposure is considered adverse may also depend on the location and its intended use
(i.e., city park, commercial, cropland). Support for consideration of adversity beyond
the species level by making explicit the linkages between stress-related effects at the
species and effects at the ecosystem level is found in A Framework for Assessing and
Reporting on Ecological Condition: an SAB report (U.S. EPA, 2002). Additionally,
the National Acid Precipitation Assessment Program (NAPAP, 1991) uses the
following working definition of "adverse ecological effects" in the preparation of
reports to Congress mandated by the Clean Air Act: "any injury (i.e., loss of
chemical or physical quality or viability) to any ecological or ecosystem component,
up to and including at the regional level, over both long and short terms."
On a broader scale, ecosystem services may provide indicators for ecological
impacts. Ecosystem services are the benefits that people obtain from ecosystems
(UNEP, 2003). According to the Millennium Ecosystem Assessment, ecosystem
services include: "provisioning services such as food and water; regulating services
such as regulation of floods, drought, land degradation, and disease; supporting
services such as soil formation and nutrient cycling; and cultural services such as
recreational, spiritual, religious, and other nonmaterial benefits." For example, a
more subtle ecological effect of pollution exposure may result in a clearly adverse
impact on ecosystem services if it results in a population decline in a species that is
recreationally or culturally important.
Ixxii
-------
References
ATS (American Thoracic Society). (2000b). What constitutes an adverse health effect of air pollution? This
official statement of the American Thoracic Society was adopted by the ATS Board of Directors, July 1999.
Am J Respir Crit Care Med 161: 665-673.
Bell. ML; Dominici. F; Samet. JM. (2005). A meta-analysis of time-series studies of ozone and mortality with
comparison to the national morbidity, mortality, and air pollution study. Epidemiology 16: 436-445.
http://dx.doi.org/10.1097/01.ede.0000165817.40152.85
CDC (Centers for Disease Control and Prevention). (2004). The health consequences of smoking: A report of
the Surgeon General. Washington, DC: U.S. Department of Health and Human Services.
http ://www. surgeongeneral.gov/librarv/smokingconsequences/
Fox. GA. (1991). Practical causal inference for ecoepidemiologists. JToxicol Environ Health A 33: 359-373.
http://dx.doi.org/10.1080/15287399109531535
Gee. GC: Pavne-Sturges. DC. (2004). Environmental health disparities: A framework integrating psychosocial
and environmental concepts. Environ Health Perspect 112: 1645-1653. http://dx.doi.org/10.1289/ehp.7074
HEW (U.S. Department of Health, Education and Welfare). (1964). Smoking and health: Report of the
advisory committee to the surgeon general of the public health service. Washington, DC: U.S. Department
of Health, Education, and Welfare. http://profiles.nlm.nih.gov/ps/retrieve/ResourceMetadata/NNBBMQ
Hill. AB. (1965). The environment and disease: Association or causation? Proc R Soc Med 58: 295-300.
IARC (International Agency for Research on Cancer). (2006). Preamble to the IARC monographs. Lyon,
France. http://monographs.iarc.fr/ENG/Preamble/
loannidis. JPA. (2008). Why most discovered true associations are inflated [Review]. Epidemiology 19: 640-
648. http://dx.doi.org/10.1097/EDE.Ob013e31818131e7
IPCC (Intergovernmental Panel on Climate Change). (2007a). Climate change 2007: Impacts, adaptation and
vulnerability. Cambridge, UK: Cambridge University Press.
NAPAP (National Acid Precipitation Assessment Program). (1991). The experience and legacy of NAPAP:
Report of the Oversight Review Board of the National Acid Precipitation Assessment Program.
Washington, DC.
Rothman. KJ; Greenland. S. (1998). Modern epidemiology (2nd ed.). Philadelphia, PA: Lippincott, Williams,
& Wilkins.
Samet. JM; Bodurow. CC. (2008). Improving the presumptive disability decision-making process for veterans.
In JM Samet; CC Bodurow (Eds.). Washington, DC: National Academies Press.
http://www.nap.edu/openbook.php7record id=l 1908
U.S. EPA (U.S. Environmental Protection Agency). (1998). Guidelines for ecological risk assessment [EPA
Report]. (EPA/630/R-95/002F). Washington, DC. http://www.epa.gov/raf/publications/guidelines-
ecological-risk-assessment.htm
U.S. EPA (U.S. Environmental Protection Agency). (2002). A framework for assessing and reporting on
ecological condition: An SAB report [EPA Report]. (EPA-SAB-EPEC-02-009). Washington, DC.
http ://www.ntis. gov/search/product. aspx?ABBR=PB2004100741
U.S. EPA (U.S. Environmental Protection Agency). (2005). Guidelines for carcinogen risk assessment [EPA
Report]. (EPA/630/P-03/001F). Washington, DC. http://www.epa.gov/cancerguidelines/
U.S. EPA (U.S. Environmental Protection Agency). (2006b). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/R-05/004AF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
Ixxiii
-------
U.S. EPA (U.S. Environmental Protection Agency). (2009d). Integrated science assessment for participate
matter [EPA Report]. (EPA/600/R-08/139F). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=216546
UNEP (United Nations Environment Programme). (2003). Ecosystems and human well-being: A framework
for assessment. Washington, DC: Island Press.
Zeger. SL; Thomas. D; Dominici. F; Samet. JM; Schwartz. J; Dockerv. D; Cohen. A. (2000). Exposure
measurement error in time-series studies of air pollution: Concepts and consequences. Environ Health
Perspect 108: 419-426.
Ixxiv
-------
LEGISLATIVE AND HISTORICAL BACKGROUND
Legislative Requirements for the NAAQS Review
Two sections of the Clean Air Act (CAA) govern the establishment and revision of
the National Ambient Air Quality Standards (NAAQS). Section 108 (42 USC §7408)
directs the Administrator to identify and list certain air pollutants and then to issue air
quality criteria for those pollutants. The Administrator is to list those air pollutants
that in her "judgement; cause or contribute to air pollution which may reasonably be
anticipated to endanger public health or welfare;" ... "the presence of which in the
ambient air results from numerous or diverse mobile or stationary sources" and "for
which ... [the Administrator] plans to issue air quality criteria..." (CAA. 1990a). Air
quality criteria are intended to "accurately reflect the latest scientific knowledge
useful in indicating the kind and extent of identifiable effects on public health or
welfare, which may be expected from the presence of [a] pollutant in ambient air ..."
[42 USC §7408(b)].
Section 109 (CAA, 1990b) directs the Administrator to propose and promulgate
"primary" and "secondary" NAAQS for pollutants for which air quality criteria have
been issued. Section 109(b)(l) defines a primary standard as one "the attainment and
maintenance of which in the judgment of the Administrator, based on such criteria
and allowing an adequate margin of safety, are requisite to protect the public
health."1 A secondary standard, as defined in section 109(b)(2), must "specify a level
of air quality the attainment and maintenance of which, in the judgment of the
Administrator, based on such criteria, is required to protect the public welfare from
any known or anticipated adverse effects associated with the presence of [the]
pollutant in the ambient air."2
The requirement that primary standards include an adequate margin of safety was
intended to address uncertainties associated with inconclusive scientific and technical
information available at the time of standard setting. It was also intended to provide a
reasonable degree of protection against hazards that research has not yet identified.
See Lead Industries Association v. EPA, 647 F.2d 1130, 1154 (D.C. Cir 1980), cert.
denied, 449 U.S. 1042 (1980); American Petroleum Institute v. Costle, 665 F.2d
1176, 1186 (D.C. Cir. (1981), cert, denied, 455 U.S. 1034 (1982). Both kinds of
uncertainties are components of the risk associated with pollution at levels below
those at which human health effects can be said to occur with reasonable scientific
certainty. Thus, in selecting primary standards that include an adequate margin of
safety, the Administrator is seeking not only to prevent pollution levels that have
been demonstrated to be harmful but also to prevent lower pollutant levels that may
1 The legislative history of section 109 indicates that a primary standard is to be set at "the maximum permissible ambient air level. .
. which will protect the health of any [sensitive] group of the population," and that for this purpose "reference should be made to a
representative sample of persons comprising the sensitive group rather than to a single person in such a group" [S. Rep. No. Si-
ll 96, 91st Cong., 2d Sess. 10(1970)].
2 Welfare effects as defined in section 302(h) include, but are not limited to, "effects on soils, water, crops, vegetation, man-made
materials, animals, wildlife, weather, visibility and climate, damage to and deterioration of property, and hazards to transportation,
as well as effects on economic values and on personal comfort and well-being" (CAA. 2005).
Ixxv
-------
pose an unacceptable risk of harm, even if the risk is not precisely identified as to
nature or degree. The CAA does not require the Administrator to establish a primary
NAAQS at a zero-risk level or at background concentration levels, see Lead
Industries v. EPA, 647 F.2d at 1156 n.51, but rather at a level that reduces risk
sufficiently so as to protect public health with an adequate margin of safety.
In addressing the requirement for a margin of safety, EPA considers such factors as
the nature and severity of the health effects involved, the size of the sensitive
population(s) at risk, and the kind and degree of the uncertainties that must be
addressed. The selection of any particular approach to providing an adequate margin
of safety is a policy choice left specifically to the Administrator's judgment. See
Lead Industries Association v. EPA, supra, 647 F.2d at 1161-1162; Whitman v.
American Trucking Associations, 531 U.S. 457, 495 (2001).
In setting standards that are "requisite" to protect public health and welfare, as
provided in Section 109(b), EPA's task is to establish standards that are neither more
nor less stringent than necessary for these purposes. In so doing, EPA may not
consider the costs of implementing the standards. [See generally, Whitman v.
American Trucking Associations, 531 U.S. 457, 465-472, 475-76. (2001)]. Likewise,
"[attainability and technological feasibility are not relevant considerations in the
promulgation of national ambient air quality standards." American Petroleum
Institute v. Costle, 665 F. 2d at 1185.
Section 109(d)(l) requires that "not later than December 31, 1980, and at 5-year
intervals thereafter, the Administrator shall complete a thorough review of the
criteria published under section 108 and the national ambient air quality standards ...
and shall make such revisions in such criteria and standards and promulgate such
new standards as may be appropriate..." Section 109(d)(2) requires that an
independent scientific review committee "shall complete a review of the criteria ...
and the national primary and secondary ambient air quality standards ... and shall
recommend to the Administrator any new ... standards and revisions of existing
criteria and standards as may be appropriate ..." Since the early 1980s, this
independent review function has been performed by CASAC.
History of the NAAQS for Ozone
Tropospheric (ground-level) O3 is the indicator for the mix of photochemical
oxidants (e.g., peroxyacetyl nitrate, hydrogen peroxide) formed from biogenic and
anthropogenic precursor emissions. Naturally occurring O3 in the troposphere can
result from biogenic organic precursors reacting with naturally occurring nitrogen
oxides (NOx) and by stratospheric O3 intrusion into the troposphere. Anthropogenic
precursors of O3, especially NOX, and volatile organic compounds (VOCs), originate
from a wide variety of stationary and mobile sources. Ambient O3 concentrations
produced by these emissions are directly affected by temperature, solar radiation,
wind speed, and other meteorological factors.
Ixxvi
-------
NAAQS are comprised of four basic elements: indicator, averaging time, level, and
form. The indicator defines the pollutant to be measured in the ambient air for the
purpose of determining compliance with the standard. The averaging time defines the
time period over which air quality measurements are to be obtained and averaged or
cumulated, considering evidence of effects associated with various time periods of
exposure. The level of a standard defines the air quality concentration used (i.e., an
ambient concentration of the indicator pollutant) in determining whether the standard
is achieved. The form of the standard specifies the air quality measurements that are
to be used for compliance purposes (e.g., the annual fourth-highest daily maximum
8-h concentration, averaged over 3 years), and whether the statistic is to be averaged
across multiple years. These four elements taken together determine the degree of
public health and welfare protection afforded by the NAAQS.
Table
Summary of primary and secondary NAAQS promulgated for O3
during the period 1971-2008.
Final Rule
1971 (36 FR 8186)
Indicator
Total
photochemical
oxidants
Avg
Time
1-h
Level
(ppm)
0.08
Form
Not to be exceeded more than 1
hour per year
1979 (44 FR 8202)
03
1-h
0.12
Attainment is defined when the expected number
of days per calendar year, with maximum hourly
average concentration greater than 0.12 ppm, is
1993 (58 FR 13008) EPA decided that revisions to the standards were not warranted at the time.
1997 (62 FR 38856)
03
8-h
0.08
Annual fourth-highest daily maximum 8-h
concentration averaged over 3 years
2008 (73 FR 16483)
03
8-h
0.075
Form of the standards remained unchanged
relative to the 1997 standard
Table III summarizes the O3 NAAQS that have been promulgated to date. In each
review, the secondary standard has been set to be identical to the primary standard.
These reviews are briefly described below.
EPA first established primary and secondary NAAQS for photochemical oxidants in
1971 . Both primary and secondary standards were set at a level of 0.08 parts per
million (ppm), 1-h avg, total photochemical oxidants, not to be exceeded more than
1 hour per year. The standards were based on scientific information contained in the
1970 O, AQCD.
In 1977, EPA announced the first periodic review of the 1970 AQCD in accordance
with Section 109(d)(l) of the Clean Air Act. In 1978, EPA published an AQCD.
Based on the 1978 AQCD, EPA published proposed revisions to the original
NAAQS in 1978 (U.S. EPA. 1978b) and final revisions in 1979 (U.S. EPA. 1979a).
The level of the primary and secondary standards was revised from 0.08 to 0.12 ppm;
Ixxvii
-------
the indicator was revised from photochemical oxidants to O3; and the form of the
standards was revised from a deterministic to a statistical form, which defined
attainment of the standards as occurring when the expected number of days per
calendar year with maximum hourly average concentration greater than 0.12 ppm is
equal to or less than one.
In 1982, EPA announced plans to revise the 1978 AQCD (U.S. EPA. 1978a).
In 1983, EPA announced that the second periodic review of the primary and
secondary standards for O3 had been initiated (U.S. EPA, 1983). EPA subsequently
published the 1986 O3 AQCD (U.S. EPA. 1986) and 1989 Staff Paper (U.S. EPA.
1989). Following publication of the 1986 O3 AQCD, a number of scientific abstracts
and articles were published that appeared to be of sufficient importance concerning
potential health and welfare effects of O3 to warrant preparation of a Supplement to
the 1986 O3 AQCD (Costa etal., 1992). Under the terms of a court order, on August
10, 1992, EPA published a proposed decision (U.S. EPA, 1992) stating that revisions
to the existing primary and secondary standards were not appropriate at the time
(U.S. EPA, 1992). This notice explained that the proposed decision would complete
EPA's review of information on health and welfare effects of O3 assembled over a
7-year period and contained in the 1986 O3 AQCD (U.S. EPA. 1986) and its
Supplement to the 1986 O3 AQCD (Costa et al., 1992). The proposal also announced
EPA's intention to proceed as rapidly as possible with the next review of the air
quality criteria and standards for O3 in light of emerging evidence of health effects
related to 6- to 8-hour O3 exposures. On March 9, 1993, EPA concluded the review
by deciding that revisions to the standards were not warranted at that time (U.S.
EPA. 1993).
In August 1992, EPA announced plans to initiate the third periodic review of the air
quality criteria and O3 NAAQS (U.S. EPA. 1992). On the basis of the scientific
evidence contained in the 1996 O3 AQCD (U.S. EPA. 1996a) and the 1996 Staff
Paper (U.S. EPA, 1996e), and related technical support documents, linking exposures
to ambient O3 to adverse health and welfare effects at levels allowed by the then
existing standards, EPA proposed to revise the primary and secondary O3 standards
on December 13, 1996 (U.S. EPA. 1996d). The EPA proposed to replace the then
existing 1-hour primary and secondary standards with 8-h avg O3 standards set at a
level of 0.08 ppm (equivalent to 0.084 ppm using standard rounding conventions).
The EPA also proposed, in the alternative, to establish a new distinct secondary
standard using a biologically based cumulative seasonal form. The EPA completed
the review on July 18,1997 by setting the primary standard at a level of 0.08 ppm,
based on the annual fourth-highest daily maximum 8-h avg concentration, averaged
over 3 years, and setting the secondary standard identical to the revised primary
standard (U.S. EPA. 1997).
On May 14, 1999, in response to challenges to EPA's 1997 decision by industry and
others, the U.S. Court of Appeals for the District of Columbia Circuit (D.C. Cir.)
remanded the O3 NAAQS to EPA, finding that Section 109 of the CAA, as
interpreted by EPA, effected an unconstitutional delegation of legislative authority.
In addition, the D.C. Cir. directed that, in responding to the remand, EPA should
Ixxviii
-------
consider the potential beneficial health effects of O3 pollution in shielding the public
from the effects of solar ultraviolet (UV) radiation, as well as adverse health effects.
On January 27, 2000, EPA petitioned the U.S. Supreme Court for certiorari on the
constitutional issue (and two other issues) but did not request review of the D.C. Cir.,
ruling regarding the potential beneficial health effects of O3. On February 27, 2001,
the U.S. Supreme Court unanimously reversed the judgment of the D.C. Cir. on the
constitutional issue, holding that Section 109 of the CAA does not delegate
legislative power to the EPA in contravention of the Constitution, and remanded the
case to the D.C. Cir. to consider challenges to the O3 NAAQS that had not been
addressed by that Court's earlier decisions. On March 26, 2002, the D.C. Cir. issued
its final decision, finding the 1997 O3 NAAQS to be "neither arbitrary nor
capricious," and denied the remaining petitions for review. On November 14, 2001,
in response to the D.C. Cir. remand to consider the potential beneficial health effects
of O3 pollution in shielding the public from effects of solar (UV) radiation, EPA
proposed to leave the 1997 8-h O3 NAAQS unchanged (U.S. EPA. 2001). After
considering public comment on the proposed decision, EPA published its final
response to this remand on January 6, 2003, reaffirming the 8-h O3 NAAQS set in
1997 (U.S. EPA. 2003). On April 30, 2004, EPA announced the decision to make the
1-h O3 NAAQS no longer applicable to areas 1 year after the effective date of the
designation of those areas for the 8-h NAAQS (U.S. EPA. 2004). For most areas, the
date that the 1-h NAAQS no longer applied was June 15, 2005.
EPA initiated the next periodic review of the air quality criteria and O3 standards in
September 2000 with a call for information (U.S. EPA, 2000). The schedule for
completion of that rulemaking later became governed by a consent decree resolving a
lawsuit filed in March 2003 by a group of plaintiffs representing national
environmental and public health organizations. Based on the 2006 O3 AQCD (U.S.
EPA. 2006b) published in March 2006, the Staff Paper (U.S. EPA. 2007b) and
related technical support documents, the proposed decision was published in the
Federal Register on July 11, 2007 (U.S. EPA. 2007a). The EPA proposed to revise
the level of the primary standard to a level within the range of 0.075 to 0.070 ppm.
Two options were proposed for the secondary standard: (1) replacing the current
standard with a cumulative, seasonal standard, expressed as an index of the annual
sum of weighted hourly concentrations cumulated over 12 daylight hours during the
consecutive 3-month period within the O3 season with the maximum index value, set
at a level within the range of 7 to 21 ppm-h; and (2) setting the secondary standard
identical to the revised primary standard. The EPA completed the rulemaking with
publication of a final decision on March 27, 2008 (U.S. EPA. 2008f). revising the
level of the 8-hour primary O3 standard from 0.08 ppm to 0.075 ppm and revising the
secondary standard to be identical to the primary standard.
In May 2008, state, public health, environmental, and industry petitioners filed suit
against EPA regarding that final decision. At EPA's request the consolidated cases
were held in abeyance pending EPA's reconsideration of the 2008 decision. A notice
of proposed rulemaking to reconsider the 2008 final decision was issued by the
Administrator on January 6, 2010. Three public hearings were held. The Agency
solicited CASAC review of the proposed rule on January 25, 2010 and additional
Ixxix
-------
CAS AC advice on January 26, 2011. On September 2, 2011, the Office of
Management and Budget returned the draft final rule on reconsideration to EPA for
further consideration. EPA decided to coordinate further proceedings on its voluntary
rulemaking on reconsideration with the ongoing periodic review, by deferring the
completion of its voluntary rulemaking on reconsideration until it completes its
statutorily-required periodic review. In light of that, the litigation on the 2008 final
decision is no longer being held in abeyance and is proceeding. The 2008 O3
standards remain in effect.
Ixxx
-------
References
CAA. Clean Air Act, as amended by Pub. L. No. 101-549. section 108: Air quality criteria and control
techniques. § 7408 (1990a). http://www.law.Cornell.edu/uscode/text/42/7408
CAA. Clean Air Act, as amended by Pub. L. No. 101-549. section 109: National primary and secondary
ambient air quality standards. § 7409 (1990b). http://www.epa.gov/air/caa/titlel.html#ia
CAA. Clean Air Act, section 302: Definitions. § 7602 (2005). http://www.gpo.gov/fdsvs/pkg/USCODE-2005-
title42/pdf/USCODE-2005-title42-chap85-subchapIII-sec7602.pdf
Costa. PL; Folinsbee. LJ; Raub. JA; Tilton. B; Tingev. DT. (1992). Summary of selected new information on
effects of ozone on health and vegetation: Supplement to 1986 air quality criteria for ozone and other
photochemical oxidants. (EPA/600/8-88/105F). Research Triangle Park, NC: U.S. Environmental
Protection Agency. http://cfpub.epa.gov/ncea/isa/recordisplav.cfm?deid=31093
U.S. EPA (U.S. Environmental Protection Agency). (1971). National primary and secondary ambient air
quality standards. Fed Reg 36: 8186-8201.
U.S. EPA (U.S. Environmental Protection Agency). (1978a). Air quality criteria for ozone and other
photochemical oxidants [EPA Report]. (EPA/600/8-78/004). Washington, DC.
U.S. EPA (U.S. Environmental Protection Agency). (1978b). Photochemical oxidants: Proposed revisions to
the national ambient air quality standards. Fed Reg 43: 26962-26971.
U.S. EPA (U.S. Environmental Protection Agency). (1979a). National primary and secondary ambient air
quality standards: Revisions to the national ambient air quality standards for photochemical oxidants. Fed
Reg 44: 8202-8237.
U.S. EPA (U.S. Environmental Protection Agency). (1982). Air quality criteria document for ozone and other
photochemical oxidants. Fed Reg 47: 11561.
U.S. EPA (U.S. Environmental Protection Agency). (1983). Review of the national ambient air quality
standards for ozone. Fed Reg 48: 38009.
U.S. EPA (U.S. Environmental Protection Agency). (1986). Air quality criteria for ozone and other
photochemical oxidants [EPA Report]. (EPA-600/8-84-020aF - EPA-600/8-84-020eF). Research Triangle
Park, NC. http://www.ntis.gov/search/product.aspx?ABBR=PB87142949
U.S. EPA (U.S. Environmental Protection Agency). (1989). Review of the national ambient air quality
standards for ozone: Assessment of scientific and technical information: OAQPS staff report [EPA Report].
(EPA-45 0/2-92-001). Research Triangle Park, NC.
http://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=2000LOW6.txt
U.S. EPA (U.S. Environmental Protection Agency). (1992). National ambient air quality standards for ozone;
Proposed decision. Fed Reg 57: 35542-35557.
U.S. EPA (U.S. Environmental Protection Agency). (1993). National ambient air quality standards for ozone -
Final decision. Fed Reg 58: 13008-13019.
U.S. EPA (U.S. Environmental Protection Agency). (1996d). National ambient air quality standards for ozone:
Proposed decision. Fed Reg 61: 65716-65750.
U.S. EPA (U.S. Environmental Protection Agency). (1996e). Review of national ambient air quality standards
for ozone: Assessment of scientific and technical information: OAQPS staff paper [EPA Report].
(EPA/452/R-96/007). Research Triangle Park, NC.
http ://www.ntis. gov/search/product. aspx?ABBR=PB9620343 5
U.S. EPA (U.S. Environmental Protection Agency). (1997). National ambient air quality standards for ozone;
final rule. Fed Reg 62: 38856-38896.
Ixxxi
-------
U.S. EPA (U.S. Environmental Protection Agency). (2000). Air quality criteria for ozone and related
photochemical oxidants. Fed Reg 65: 57810.
U.S. EPA (U.S. Environmental Protection Agency). (2001). National ambient air quality standards for ozone:
Proposed response to remand. Fed Reg 66: 57268-57292.
U.S. EPA (U.S. Environmental Protection Agency). (2003). National ambient air quality standards for ozone:
Final response to remand. Fed Reg 68: 614-645.
U.S. EPA (U.S. Environmental Protection Agency). (2004). Final rule to implement the 8-hour ozone national
ambient air quality standard-phase 1. Fed Reg 69: 23951-24000.
U.S. EPA (U.S. Environmental Protection Agency). (2006b). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/R-05/004AF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
U.S. EPA (U.S. Environmental Protection Agency). (2007a). National ambient air quality standards for ozone.
Fed Reg 72: 37818-37919.
U.S. EPA (U.S. Environmental Protection Agency). (2007b). Review of the national ambient air quality
standards for ozone: Policy assessment of scientific and technical information: OAQPS staff paper [EPA
Report]. (EPA/452/R-07/003). Research Triangle Park, NC.
http://www.epa.gov/ttn/naaqs/standards/ozone/data/2007 01 ozone staff_paper
U.S. EPA (U.S. Environmental Protection Agency). (2008f). National ambient air quality standards for ozone.
Fed Reg 73: 16436-16514.
Ixxxii
-------
1 EXECUTIVE SUMMARY
Introduction and Purpose
The purpose of this Integrated Science Assessment (ISA) is to provide a synthesis
and evaluation of the most policy-relevant science that builds the scientific
foundation for the periodic review of the primary (health-based) and secondary
(welfare-based) national ambient air quality standard (NAAQS) for ozone (O3) and
related photochemical oxidants required by the Clean Air Act. The primary NAAQS
protects against respiratory health effects incurred after short-term exposure to O3,
while the secondary NAAQS protects against damage to vegetation and ecosystems,
generally referred to as "welfare effects." The ISA is intended to inform both EPA's
Risk and Exposure Assessment and Policy Assessment, and decisions by the EPA
Administrator on the NAAQS for O3 (see Figure I in Preamble). Set in 2008, the
current primary O3 standard is an 8-hour average standard of 75 parts per billion
(ppb). The secondary standard for O3 is equal to the primary standard.
Scope and Methods
EPA has an extensive, robust process for evaluating the latest scientific evidence and
drawing conclusions regarding air pollution-related health and welfare effects.
Building upon the findings of previous assessments, this review includes
identification, selection, evaluation, and integration of the relevant results pertaining
to the atmospheric science of O3; short- and long-term exposure to ambient O3;
health effects due to relevant O3 concentrations as characterized in epidemiologic or
controlled human exposure and toxicological studies; and ecological and other
welfare effects. Additionally, this review will characterize O3 concentration-response
relationships, mode(s) of action, and populations at increased risk for O3-related
health effects. The conclusions and key findings from previous reviews provide the
basis for the consideration of evidence from recent studies (i.e., studies published
since the completion of the 2006 O3 Air Quality Criteria Document [AQCD]).
Conclusions are drawn based on the synthesis of evidence across scientific
disciplines.
EPA assesses the body of peer-reviewed literature to draw conclusions on the causal
relationships between relevant pollutant concentrations and health or welfare effects.
EPA specifically evaluates the quantitative evidence and draws scientific
conclusions, to the extent possible, regarding the concentration-response
relationships and the loads to ecosystems, exposure doses or concentrations, and
duration and pattern of exposures at which effects are observed.
EPA uses a consistent and transparent approach to evaluate the causal nature of air
pollution-related health and environmental effects for use in developing ISAs; the
framework for causal determinations is described in the Preamble to this document.
Causality determinations are based on the evaluation and synthesis of evidence
1-1
-------
across scientific disciplines. A five-level hierarchy is used to classify the weight of
evidence for causation, not just association. This weight of evidence evaluation is
based on various lines of evidence from across the health and environmental effects
disciplines. These separate conclusions are integrated into a qualitative statement
about the overall weight of the evidence and causality. The causal determinations are:
• Causal relationship
• Likely to be a causal relationship
• Suggestive of a causal relationship
• Inadequate to infer a causal relationship
• Not likely to be a causal relationship
Ambient Ozone Concentrations
Ozone is naturally present in the stratosphere (an elevated layer of the Earth's
atmospehre) where it serves the beneficial role of absorbing harmful ultraviolet
radiation from the Sun, preventing the majority of this radiation from reaching the
surface of the Earth. However, in the troposphere (the layer of the atmosphere
extending from the stratosphere down to the Earth's surface), O3 acts as a powerful
oxidizing agent, which can harm living organisms and materials. Tropospheric O3 is
present not only in polluted urban air, but across the globe. Ozone concentrations can
be influenced by local meteorological conditions, circulation patterns, emissions, and
topographic barriers, resulting in heterogeneous concentrations across an individual
urban area. On a larger scale, O3 can last in the atmosphere long enough that it can
be transported from continent to continent.
Ozone in the troposphere originates from both anthropogenic (i.e., man-made) and
natural sources. Ozone attributed to anthropogenic sources is formed by
photochemical reactions involving sunlight and precursor pollutants, including
volatile organic compounds (VOCs), nitrogen oxides (NOX), and carbon monoxide
(CO). Ozone attributed to natural sources is formed through the same photochemical
reactions involving natural emissions of precursor pollutants from vegetation,
microbes, animals, burning biomass (e.g., forest fires), and lightning. Ozone is lost
through deposition to surfaces and chemical reactions occurring in the atmosphere.
The highest O3 concentrations are not found in urban areas close to the concentrated
sources of its precursors such as traffic, but rather in suburban and rural areas
downwind of these sources. Reaction of O3 with NO in fresh motor vehicle exhaust
depletes O3 in urban cores; but O3 can be regenerated during transport downwind of
urban source areas. Also O3 tends to be more uniform in rural than in urban areas
because O3 production occurs over large areas and because the concentrated sources
of NO depleting O3 in urban cores are generally lacking (except near power plants
and other strong sources of NO).
In the context of a review of the NAAQS, it is useful to define background O3
concentrations in a way that distinguishes between concentrations that result from
precursor emissions that are relatively less controllable (including natural sources
1-2
-------
and foreign anthropogenic sources) from those that are relatively more controllable
through U.S. policies. For example, U.S. background O3 concentrations can be
defined as those concentrations resulting from natural sources everywhere in the
world plus anthropogenic sources outside the U.S.; North American (NA)
background O3 concentrations can be defined as those concentrations resulting from
natural sources everywhere in the world plus anthropogenic sources outside the U.S.,
Canada and Mexico. Since background defined either of these ways is a hypothetical
construct that cannot be measured, NA background O3 concentrations are estimated
using chemistry transport models (see Sections 2.2 and 3.4).
Human Exposure to Ozone
The widespread presence of O3 in the environment results in exposure as people
participate in normal daily activities. Exposure may occur indoors, where people
spend most of their time, as well as outdoors, where O3 concentrations are highest.
Evaluating relationships among outdoor concentration, indoor concentration, and
personal exposure is useful for determining how well ambient monitors represent
exposure. The relationship between personal exposure and ambient concentration
measured at fixed-site monitors can be described in terms of correlation (how they
covary over time) and ratio, which describes their relative magnitude. High
correlations imply that changes in ambient concentrations are reflected in personal
exposure, while high ratios imply that the magnitude of exposure is similar to the
magnitude of concentrations. Personal-ambient O3 correlations are generally
moderate (0.3-0.8), although low correlations have been observed with increased
time spent indoors, low indoor-outdoor air exchange rates, and concentrations below
personal sampler detection limits (see Section 4.3). Ratios of 0.1-0.3 between
personal exposure and ambient concentration have been observed for the general
population, with ratios of up to 0.9 observed for outdoor workers. Evidence suggests
that some groups, particularly children, older adults, and those with respiratory
problems, change their behavior on high-O3 days when O3 warnings are issued with
advice to reduce exposure (see Section 4.4.2). Such behavioral changes may result in
reduced effect estimates in epidemiologic studies that do not account for averting
behavior on high-O3 days. Variation in O3 concentrations occurs over multiple
spatial and temporal scales, and this introduces exposure error into epidemiologic
results (see Section 4.6.2). However, epidemiologic studies evaluating the influence
of spatial scale and monitor selection find little difference among effect estimates,
and comparable risk estimates have been reported in studies using a variety of
exposure assessment techniques expected to produce different levels of personal-
ambient associations. This suggests that there is no clear indication that any
particular method of exposure assessment produces stronger epidemiologic results
than any other method.
1-3
-------
Dosimetry and Modes of Action
When O3 is inhaled, the amount of O3 that is absorbed is affected by the shape and
size of the respiratory tract, and the route of breathing (nose or mouth), as well as
how quickly and deeply a person is breathing. The amount of O3 that is removed
from the air stream during breathing is referred to as uptake. The primary site of O3
uptake moves deeper into the respiratory tract during exercise when breathing
becomes faster and the breathing route changes from the nose only to oronasal
breathing (i.e., through the nose and mouth) (see Section 5.2). Tissue dose refers to
the amount of O3 that diffuses through the lining of the lung and reaches the
underlying tissues. The site of the greatest O3 dose to the lung tissue is the junction
of the conducting airway and the gas exchange region in the deeper portion of the
lung.
Once O3 has been absorbed, there are several key events in the toxicity pathway of
O3 in the respiratory tract that lead to O3-induced health effects (see Section 5.3).
These include formation of secondary oxidation products in the lung, activation of
neural reflexes, initiation of inflammation, alterations of epithelial barrier function,
sensitization of bronchial smooth muscle, modification of innate and adaptive
immunity, and airway remodeling. Another key event, systemic inflammation and
vascular oxidative/nitrosative stress, may be critical to the extrapulmonary effects of
03.
Overall, biological responses to O3 exposure are similar across many species (see
Section 5.5). Thus, animal studies are used to add to the understanding of the full
range of potential O3-mediated health effects.
Integration of Ozone Health Effects
In this ISA, the body of evidence from short-term (i.e., hours, days, weeks) or long-
term (i.e., months to years) exposure studies is evaluated and integrated across
relevant scientific disciplines (i.e., controlled human exposure studies, toxicology,
and epidemiology) for health effects that vary in severity from minor subclinical
effects to death. The results from the health studies, supported by the evidence from
atmospheric chemistry and exposure assessment studies, contribute to the causal
determinations made for the various health outcomes. Both the conclusions from the
2006 O3 AQCD and the causality determinations from this review are summarized in
Table 1-1. Additional details are provided here for respiratory and cardiovascular
health effects and mortality, for which there is the strongest evidence of an effect
from O3; details for a wider range of health effects are provided in subsequent
chapters1.
1 Detailed information O3 concentrations at which health effects are observed can be found in later chapters. For an overview,
please see Table 2-2.
1-4
-------
Table 1-1 Summary of O3 causal determinations by exposure duration and
health outcome.
Health Outcome3
Conclusions from 2006 O3 AQCD
Conclusions from this ISA
Short-term Exposure to
Respiratory effects
The overall evidence supports a causal
relationship between acute ambient Os
exposures and increased respiratory morbidity
outcomes.
Causal Relationship
Cardiovascular
effects
The limited evidence is highly suggestive that
Os directly and/or indirectly contributes to
cardiovascular-related morbidity, but much
remains to be done to more fully substantiate
the association.
Likely to be a
Causal Relationship
Central nervous Toxicological studies report that acute
system effects exposures to Os are associated with alterations
in neurotransmitters, motor activity, short and
long term memory, sleep patterns, and
histological signs of neurodegeneration.
Suggestive of a
Causal Relationship
Total Mortality
The evidence is highly suggestive that Os
directly or indirectly contributes to non-
accidental and cardiopulmonary-related
mortality.
Likely to be a
Causal Relationship
Long-term Exposure to O3
Respiratory effects
The current evidence is suggestive but
inconclusive for respiratory health effects from
long-term Os exposure.
Likely to be a
Causal Relationship
Cardiovascular
effects
No conclusions in the 2006 O3 AQCD.
Suggestive of a
Causal Relationship
Reproductive and
developmental
effects
Limited evidence for a relationship between air
pollution and birth-related health outcomes,
including mortality, premature births, low birth
weights, and birth defects, with little evidence
being found for Os effects.
Suggestive of a
Causal Relationship
Central nervous Evidence regarding chronic exposure and
system effects neurobehavioral effects was not available.
Suggestive of a
Causal Relationship
Cancer
Little evidence for a relationship between
chronic O3 exposure and increased risk of lung
cancer.
Inadequate to Infer a
Causal Relationship
Total Mortality
There is little evidence to suggest a causal
relationship between chronic O3 exposure and
increased risk for mortality in humans.
Suggestive of a
Causal Relationship
aHealth effects (e.g., respiratory effects, cardiovascular effects) include a spectrum of outcomes, from measureable
subclinical effects (e.g., blood pressure), to more obvious effects (e.g., medication use, hospital admissions), and cause-
specific mortality. Total mortality includes all-cause (non-accidental) mortality, as well as cause-specific mortality
(e.g., deaths due to heart attacks).
1-5
-------
Respiratory Effects
The clearest evidence for health effects associated with exposure to O3 is provided
by studies of respiratory effects. Collectively, a very large amount of evidence
spanning several decades supports a relationship between exposure to O3 and a broad
range of respiratory effects (see Section 6.2.9 and Section 7.2.8). The majority of this
evidence is derived from studies investigating short-term exposure (i.e., hours to
weeks) to O3, although animal toxicological studies and recent epidemiologic
evidence demonstrate that long-term exposure (i.e., months to years) may also harm
the respiratory system.
The 2006 O3 AQCD concluded that there was clear, consistent evidence of a causal
relationship between short-term exposure to O3 and respiratory health effects. This
causal association is more substantiated now by the effects observed across recent
controlled human exposure, epidemiologic, and toxicological studies indicating
associations between short-term O3 exposures and a range of respiratory health
endpoints from respiratory tract inflammation to respiratory-related emergency
department (ED) visits and hospital admissions. Short-term O3 exposures induced (or
were associated with) statistically significant declines in lung function. An equally
strong body of evidence from controlled human exposure and toxicological studies
demonstrated reversible O3-induced increases in inflammatory responses, epithelial
permeability, and airway hyperresponsiveness that were found to last for 18-24 hours
after O3 exposure. Toxicological studies in animals provided additional evidence for
O3-induced impairment of host defenses. Combined, these findings from
experimental studies provided support for epidemiologic evidence, in which short-
term increases in O3 concentration were consistently associated with increases in
respiratory symptoms and asthma medication use in children with asthma,
respiratory-related hospital admissions, and ED visits for chronic obstructive
pulmonary disease (COPD) and asthma. Additionally, recent epidemiologic evidence
supports the range of respiratory effects induced by O3 by demonstrating that short-
term increases in ambient O3 concentrations can lead to respiratory mortality.
The combined evidence from these disciplines supports the conclusion that there is a
causal relationship between short-term O3 exposure and respiratory effects.
Epidemiologic evidence for a relationship between long-term O3 exposure and
respiratory effects includes recent studies that evaluate the associations between
long-term exposure to O3 and respiratory effects that demonstrate interactions
between exercise or different genetic variants and both new-onset asthma in children
and increased respiratory symptom effects in individuals with asthma. Additional
studies of respiratory health effects and a study of respiratory mortality provide a
collective body of evidence supporting this relationship. Studies evaluating other
pollutants provide data suggesting that the effects related to O3 are independent from
the effects of the other pollutants. Short-term studies provide supportive evidence
with increases in respiratory symptoms and asthma medication use, hospital
admissions and ED visits for all respiratory outcomes and asthma, and decreased
lung function in children. Taken together, the recent epidemiologic studies of
1-6
-------
respiratory health effects (including symptoms, new-onset asthma and mortality)
combined with toxicological studies in rodents and nonhuman primates, provide
biologically plausible evidence that there is likely to be a causal relationship
between long-term exposure to O3 and respiratory effects.
Mortality Effects
The last review concluded that the overall body of evidence was highly suggestive
that short-term exposure to O3 directly or indirectly contributes to non-accidental and
cardiopulmonary-related mortality, but that additional research was needed to more
fully establish the underlying mechanisms by which such effects occur. Recent
multicity studies and a multicontinent study have reported associations between
short-term O3 exposure and mortality, expanding upon evidence available in the last
review (see Section 6.6). These recent studies reported consistent positive
associations between short-term O3 exposure and total (nonaccidental) mortality,
with associations being stronger during the warm season, when O3 concentrations
were higher. They also observed associations between O3 exposure and
cardiovascular and respiratory mortality. These recent studies also examined
previously identified areas of uncertainty in the O3-mortality relationship, and
provided additional evidence supporting an association between short-term O3
exposure and mortality. As a result, the current body of evidence indicates that there
is likely to be a causal relationship between short-term exposures to O3 and
total mortality.
Cardiovascular Effects
In previous O3 reviews, very few studies were available which examined the effect
of short-term O3 exposure on the cardiovascular system. New toxicological studies,
although limited in number, have provided evidence of O3-induced cardiovascular
effects. These effects may, in part, correspond to changes in the autonomic nervous
system or to the development and maintenance of oxidative stress and inflammation
throughout the body that resulted from inflammation in the lungs. Controlled human
exposure studies also suggest cardiovascular effects in response to short-term O3
exposure, including changes in heart rate variability and blood markers of systemic
inflammation and oxidative stress, which provide some coherence with the effects
observed in animal toxicology studies. Collectively, the experimental studies provide
initial biological plausibility for the consistently positive associations observed in
epidemiologic studies of short-term O3 exposure and cardiovascular mortality.
However, studies in the epidemiologic literature generally have not observed a
relationship between short-term exposure to O3 and cardiovascular morbidity
including studies that examined the association between short-term O3 exposure and
cardiovascular-related hospital admissions and ED visits and other various
cardiovascular effects. The lack of coherence between the results from studies that
examined associations between short-term O3 exposure and cardiovascular morbidity
and cardiovascular mortality complicate the interpretation of the overall evidence for
1-7
-------
O3-induced cardiovascular effects. Although there is a lack of coherence with
epidemiologic studies of cardiovascular morbidity, animal toxicological studies
demonstrate O3-induced cardiovascular effects, and provide support to the strong
body of evidence indicating O3-induced cardiovascular mortality. Overall, the body
of evidence indicates that there is likely to be a causal relationship between
short-term exposures to O3 and cardiovascular effects, including
cardiovascular mortality.
Populations Potentially at Increased Risk
The examination of populations and lifestages potentially at increased risk for O3
exposure identifies populations that are at increased risk for O3-related health effects;
these studies do so by examining groups within the study population, such as those
with an underlying health condition or genetic variant, categories of age, race, or sex,
or by developing animal models that mimic the conditions associated with a health
effect. Such studies have identified a multitude of factors that could potentially
contribute to whether a population is at increased risk for O3-related health effects
(see Chapter 8). The populations and lifestages identified as having adequate
evidence for being at increased risk of O3-related health effects are individuals with
asthma, younger and older age groups, individuals with reduced intake of certain
nutrients (i.e., Vitamins C and E), outdoor workers, and individuals having variants
(including variations in multiple genes related to oxidative metabolism or
inflammation). The evidence for other potential factors; sex, socioeconomic status,
and obesity, is suggestive of an increased risk, but further evidence is needed.
Integration of Effects on Vegetation and Ecosystems
The most policy-relevant information pertaining to the review of the NAAQS for the
effects of O3 on vegetation and ecosystems has been evaluated and synthesized,
integrating key findings about plant physiology, biochemistry, whole plant biology,
ecosystems and exposure-response relationships. The welfare effects of O3 can be
observed across spatial scales, starting at the subcellular and cellular level, then the
whole plant and finally, ecosystem-level processes. Ozone effects at small spatial
scales, such as the leaf of an individual plant, can result in effects along a continuum
of larger spatial scales. These effects include altered rates of leaf gas exchange,
growth, and reproduction at the individual plant level, and can result in broad
changes in ecosystems, such as productivity, carbon storage, water cycling, nutrient
cycling, and community composition. The conclusions from the previous NAAQS
review and the causality determinations from this review are summarized in
Table 1-2 below. Further discussion of these is provided after Table 1-2 for: visible
foliar injury; growth, productivity, and carbon storage; yield and quality of
agricultural crops; water cycling; below-ground processes; community composition;
and O3 exposure-response relationships. Discussion of all relevant welfare effects is
provided in Chapter 9.
1-8
-------
Table 1-2 Summary of Os causal determination for welfare effects.
Vegetation and
Ecosystem Effects
Conclusions from 2006 O3 AQCD
Conclusions from
this ISA
Visible Foliar Injury
Effects on Vegetation
Data published since the 1996 O3 AQCD
strengthen previous conclusions that there is
strong evidence that current ambient Os
concentrations cause impaired aesthetic quality
of many native plants and trees by increasing foliar
injury.
Causal Relationship
Reduced Vegetation
Growth
Data published since the 1996 O3 AQCD
strengthen previous conclusions that there is
strong evidence that current ambient Os
concentrations cause decreased growth and
biomass accumulation in annual, perennial and
woody plants, including agronomic crops, annuals,
shrubs, grasses, and trees.
Causal Relationship
Reduced Productivity in
Terrestrial Ecosystems
There is evidence that O3 is an important
stressor of ecosystems and that the effects
on individual plants and processes are scaled up
through the ecosystem, affecting net primary
productivity.
Causal Relationship
Reduced Carbon (C)
Sequestration in
Terrestrial Ecosystems
Limited studies from the 2006 O3 AQCD.
Likely to be a
Causal Relationship
Reduced Yield and
Quality of Agricultural
Crops
Data published since the 1996 O3 AQCD
strengthen previous conclusions that there is
strong evidence that current ambient Os
concentrations cause decreased yield and/or
nutritive quality in a large number of agronomic
and forage crops.
Causal Relationship
Alteration of Terrestrial
Ecosystem Water
Cycling
Ecosystem water quantity may be affected by Os
exposure at the landscape level.
Likely to be a
Causal Relationship
Alteration of Below-
ground Biogeochemical
Cycles
Ozone-sensitive species have well known
responses to O3 exposure, including altered
carbon (C) allocation to below-ground tissues, and
also altered rates of leaf and root production,
turnover, and decomposition. These shifts can
affect overall C and nitrogen (N) loss from the
ecosystem in terms of respired C, and leached
aqueous dissolved organic and inorganic C and N.
Causal Relationship
Alteration of Terrestrial
Community Composition
Ozone may be affecting above- and below -ground
community composition through impacts on both
growth and reproduction. Significant changes in
plant community composition resulting directly
from O3 exposure have been demonstrated.
Likely to be a
Causal Relationship
1-9
-------
Visible Foliar Injury
Visible foliar injury resulting from exposure to O3 has been well characterized and
documented over several decades on many tree, shrub, herbaceous and crop species.
In addition, O3-induced visible foliar injury symptoms on certain plant species (e.g.,
black cherry, yellow-poplar, and common milkweed, among others) are considered
diagnostic of exposure to O3, as experimental evidence has clearly established a
consistent association, with greater exposure generally resulting in greater and more
prevalent injury. Additional sensitive species showing visible foliar injury continue
to be identified from field surveys and verified in controlled exposure studies (see
Section 9.4.2). Overall, evidence is sufficient to conclude that there is a causal
relationship between ambient O3 exposure and the occurrence of O3-induced
visible foliar injury on sensitive vegetation across the U.S.
Growth, Productivity, Carbon Storage and Agriculture
Ambient O3 concentrations have long been known to cause decreases in
photosynthetic rates and plant growth. The O3-induced effects at the plant scale may
translate to the ecosystem scale, with changes in productivity and carbon (C) storage.
Studies demonstrating the effects of O3 exposure on photosynthesis, growth, biomass
allocation, ecosystem production and ecosystem C sequestration were reviewed for
natural ecosystems (see Section 9.4.3), and crop productivity and crop quality were
reviewed for agricultural ecosystems (see Section 9.4.4). There is strong and
consistent evidence that current ambient concentrations of O3 decrease plant
photosynthesis and growth in numerous plant species across the U.S.
Studies conducted during the past four decades have also demonstrated
unequivocally that O3 alters biomass allocation and plant reproduction. Studies at the
leaf and plant scales showed that O3 reduced photosynthesis and plant growth,
providing coherence and biological plausibility for the reported decreases in
ecosystem productivity. In addition to primary productivity, other indicators such as
net ecosystem CO2 exchange and C sequestration were often assessed by modeling
studies. Model simulations consistently found that O3 exposure caused negative
impacts on those indicators, but the severity of these impacts was influenced by
multiple interactions of biological and environmental factors. Although O3 generally
causes negative effects on ecosystem productivity, the magnitude of the response
varies among plant communities. Overall, evidence is sufficient to conclude that
there is a causal relationship between ambient O3 exposure and reduced native
plant growth and productivity, and that there is a likely causal relationship
between O3 exposure and reduced carbon sequestration in terrestrial
ecosystems.
The detrimental effect of O3 on crop production has been recognized since the 1960s,
and current O3 concentrations in many areas across the U.S. are high enough to cause
yield loss in a variety of agricultural crops including, but not limited to, soybeans,
wheat, potatoes, watermelons, beans, turnips, onions, lettuces, and tomatoes.
1-10
-------
Continued increases in O3 concentration may further decrease yield in these sensitive
crops while also causing yield losses in less sensitive crops. Research has linked
increasing O3 concentration to decreased photosynthetic rates and accelerated aging
in leaves, which are related to yield (see Section 9.4.4). Recent research has
highlighted the effects of O3 on crop quality. Increasing O3 concentration decreases
nutritive quality of grasses, decreases macro- and micro-nutrient concentrations in
fruits and vegetable crops, and decreases cotton fiber quality. Evidence is sufficient
to conclude that there is a causal relationship between O3 exposure and reduced
yield and quality of agricultural crops.
Water Cycling
Ozone can affect water use in plants and ecosystems through several mechanisms,
including damage to stomatal functioning and loss of leaf area. Possible mechanisms
for O3 exposure effects on stomatal functioning include the build-up of CO2 in the
substomatal cavity, impacts on signal transduction pathways, and direct O3 impact on
guard cells. Regardless of the mechanism, O3 exposure has been shown to alter
stomatal performance, which can affect plant and stand transpiration and therefore
has the potential to affect hydrological cycling (see Section 9.4.5). Although the
direction of the response differed among studies, the evidence is sufficient to
conclude that there is likely to be a causal relationship between O3 exposure and
the alteration of ecosystem water cycling.
Below Ground Processes
Below-ground processes are tightly linked with above-ground processes.
The responses of above-ground process to O3 exposure, such as reduced
photosynthetic rates, increased metabolic cost, and reduced allocation of C to the
roots, have provided biologically plausible mechanisms for the alteration of below-
ground processes. These include altered quality and quantity of C input to soil,
changes in microbial community composition, and effects on C and nutrient cycling
(see Section 9.4.6). The evidence is sufficient to conclude that there is a causal
relationship between O3 exposure and the alteration of below-ground
biogeochemical cycles.
Community Composition
Ozone exposure changes competitive interactions and leads to loss of O3-sensitive
species or genotypes. Studies at the plant level found that the severity of O3 damage
to growth, reproduction, and foliar injury varied among species, which provide the
biological plausibility for the alteration of community composition (see Section 9.4.3
and Section 9.4.7). For example, there is a tendency for O3 exposure to shift the
biomass of grass-legume mixtures in favor of grass species. In addition, research
since the last review has shown that O3 can also alter community composition and
diversity of soil microbial communities. Shifts in community composition of bacteria
1-11
-------
and fungi have been observed in both natural and agricultural ecosystems, although
no general patterns could be identified. The evidence is sufficient to conclude that
there is likely to be a causal relationship between O3 exposure and the
alteration of community composition of some ecosystems.
Air Quality Indices and Exposure-Response
Exposure indices are metrics that quantify exposure as it relates to measured plant
response (e.g., reduced growth). They are summary measures of monitored O3
concentrations over time intended to provide a consistent metric for reviewing and
comparing exposure-response effects obtained from various studies. Given the
current state of knowledge and the best available data, exposure indices that cumulate
and differentially weight the higher hourly average concentrations and also include
the mid-level values (e.g., the W126 metric, see Section 9.5) continue to offer the
most defensible approach for use in developing response functions and comparing
studies, as well as for defining future indices for vegetation protection.
Previous reviews of the NAAQs have included exposure-response functions for the
yield of many crop species, and for the biomass accumulation of tree species. They
were based on large-scale experiments designed to obtain clear exposure-response
data, and are updated by using the W126 metric to quantify exposure. In recent years,
extensive exposure-response data obtained in more naturalistic settings have become
available for yield of soybean and growth of aspen. The exposure-response median
functions are validated based on previous data by comparing their predictions with
the newer observations (see Section 9.6). The functions supply very accurate
predictions of effects in naturalistic settings. Recent meta-analyses of large sets of
crop and tree studies do not give rise to exposure-response functions, but their results
are consistent with the functions presented in Section 9.6. Although these median
functions provide reliable models for groups of species or group of genotypes within
a species, the original data and recent results consistently show that some species,
and some genotypes within species are much more severely affected by exposure to
03.
The Role of Tropospheric Ozone in Climate Change and
UV-B Shielding Effects
Atmospheric O3 as a whole plays an important role in the Earth's energy budget by
interacting with incoming solar radiation and outgoing infrared radiation. Though
tropospheric O3 makes up only a small portion of the total amount of O3 in the
atmosphere, it has important incremental effects on the overall radiation budget.
Perturbations to tropospheric O3 concentrations can have direct effects on climate
and indirect effects on health, ecology, and welfare by changing the shielding of the
earth's surface from solar ultraviolet (UV) radiation.
1-12
-------
Radiative Forcing and Climate Change
Radiative forcing by a greenhouse gas or aerosol is a metric used to quantify the
change in balance between radiation coming into and going out of the atmosphere
caused by the presence of that substance. Tropospheric O3 is a major greenhouse gas
and radiative forcing agent; evidence from satellite data shows a sharp dip in the
outgoing infrared radiation in the 9.6 |j,m O3 absorption band. Models calculate that
the global average concentration of tropospheric O3 has doubled since the
pre-industrial era, while observations indicate that in some regions O3 may have
increased by factors as great as 4 or 5. These increases are tied to the rise in
emissions of O3 precursors from human activity, mainly fossil fuel consumption and
agricultural processes. Overall, the evidence supports a causal relationship
between changes in tropospheric O3 concentrations and radiative forcing.
The impact of the tropospheric O3 change since pre-industrial times on climate has
been estimated to be about 25-40% of the anthropogenic CO2 impact and about 75%
of the anthropogenic CH4 impact according to the Intergovernmental Panel on
Climate Change (IPCC), ranking it third in importance after CO2 and CH4 according
to the IPCC (see Section 10.3). There are large uncertainties in the magnitude of the
radiative forcing estimate attributed to tropospheric O3, making the effect of
tropospheric O3 on climate more uncertain than the effect of the longer-lived
greenhouse gases. Furthermore, radiative forcing does not take into account climate
feedbacks that could amplify or dampen the actual surface temperature response.
Quantifying the change in surface temperature requires a complex climate simulation
in which all important feedbacks and interactions are accounted for. The modeled
surface temperature response to a given radiative forcing is highly uncertain and can
vary greatly among models and from region to region within the same model. Even
with these uncertainties, global climate models indicate that tropospheric O3 has
contributed to observed changes in global mean and regional surface temperatures.
As a result of such evidence presented in climate modeling studies, there is likely to
be a causal relationship between changes in tropospheric O3 concentrations
and effects on climate.
UV-B Shielding Effects
UV radiation emitted from the Sun contains sufficient energy when it reaches the
Earth to have damaging effects on living organisms and materials (see Section 10.4).
Atmospheric O3 plays a crucial role in reducing the amount of UV radiation reaching
the Earth's surface. Ozone in the stratosphere is responsible for the majority of this
shielding, but O3 in the troposphere provides supplemental shielding of UV-B
radiation in the mid-wavelength band (280-315 nm), thereby potentially reducing
UV-B related human and ecosystem health effects and materials damage. EPA has
found no published studies that adequately examine the incremental health or welfare
effects (adverse or beneficial) attributable specifically to changes in UV-B exposure
resulting from perturbations in tropospheric O3 concentrations. While the effects are
expected to be small, they cannot yet be critically assessed within reasonable
1-13
-------
uncertainty. Overall, the evidence is inadequate to determine if a causal
relationship exists between changes in tropospheric O3 concentrations and
effects on health and welfare related to UV-B shielding.
The conclusions from the previous NAAQS review and the causality determinations
from this review relating climate change and UV-B shielding effects are summarized
in the table below (Table 1-3). with details provided in Chapter 10.
Table 1-3 Summary of O3 causal determination for climate change and UV-B
shielding effects.
Effects
Conclusions from 2006 O3 AQCD
Conclusions from this ISA
Radiative Forcing
The 2006 O3 AQCD concluded that climate forcing by O3 at
the regional scale may be its most important impact on
climate.
Causal Relationship
Climate Change
While more certain estimates of the overall importance of
global-scale forcing due to tropospheric O3 await further
advances in monitoring and chemical transport modeling,
the overall body of scientific evidence reviewed in the 2006
O3 AQCD suggests that high concentrations of Os on the
regional scale could have a discernible influence on
climate, leading to surface temperature and hydrological
cycle changes.
Likely to be a
Causal Relationship
Health and Welfare
Effects Related to
UV-B Shielding
UV-B has not been studied in sufficient detail to allow for
a credible health benefits assessment. In conclusion, the
effect of changes in surface-level Os concentrations on
UV-induced health outcomes cannot yet be critically
assessed within reasonable certainty.
Inadequate to Determine if a
Causal Relationship Exists
Conclusion
The clearest evidence for human health effects associated with exposure to O3 is
provided by studies of respiratory effects. Collectively, there is a very large amount
of evidence spanning several decades in support of a causal association between
exposure to O3 and a broad range of respiratory effects, indicating that there is a
causal relationship between short-term exposures to O3 and respiratory
effects. The majority of this evidence is derived from studies investigating short-
term O3 exposure (i.e., hours to weeks), although animal toxicological studies and
recent epidemiologic evidence demonstrate that long-term exposure (i.e., months to
years) are likely to be detrimental to the respiratory system. Additionally, consistent
positive associations between short-term O3 exposure and total (nonaccidental)
mortality have helped to resolve previously identified areas of uncertainty in the
O3-mortality relationship, indicating that there is likely to be a causal relationship
between short-term exposures to O3 and total mortality. Taken together, the
recent epidemiologic studies of respiratory health effects (including respiratory
symptoms, new-onset asthma and respiratory mortality) combined with toxicological
1-14
-------
studies in rodents and nonhuman primates, provide biologically plausible evidence
that there is likely to be a causal relationship between long-term exposure to O3
and respiratory effects. Animal toxicological studies demonstrate O3-induced
cardiovascular effects, and provide support to the strong body of evidence indicating
Os-induced cardiovascular mortality, which together indicate that there is likely to
be a causal relationship between short-term exposure to O3 and
cardiovascular effects. Recent evidence is suggestive of a causal relationship
between long-term O3 exposures and total mortality. The evidence for these
health effects indicates that the relationship between concentration and response is
linear along the range of O3 concentrations observed in the U.S., with no indication
of a threshold within that range. However, there is less certainty in the shape of the
concentration-response curve at O3 concentrations generally below 20 ppb.
The populations identified as having increased risk of O3-related health effects are
individuals with asthma, younger and older age groups, individuals with certain
dietary deficiencies, and outdoor workers.
There has been over 40 years of research on the effects of O3 exposure on vegetation
and ecosystems. The best evidence for effects is from controlled exposure studies.
These studies have clearly shown that exposure to O3 is causally linked to visible
foliar injury, decreased photosynthesis, changes in reproduction, and
decreased growth. Recently, studies at larger spatial scales support the results from
controlled studies and indicate that ambient O3 exposures can affect ecosystem
productivity, crop yield, water cycling, and ecosystem community composition. And
on a global scale, tropospheric O3 is the third most important greenhouse gas,
making it likely to play an important role in climate change.
1-15
-------
2 INTEGRATIVE SUMMARY
This Integrated Science Assessment (ISA) forms the scientific foundation for the
review of the current national ambient air quality standards (NAAQS) for ozone
(O3). The ISA is a concise evaluation and synthesis of the most policy-relevant
science—and it communicates critical science judgments relevant to the review of
the NAAQS for O3. The ISA accurately reflects "the latest scientific knowledge
useful in indicating the kind and extent of identifiable effects on public health or
welfare which may be expected from the presence of [a] pollutant in ambient air"
(CAA, 1990a). Key information and judgments contained in prior Air Quality
Criteria Documents (AQCD) for O3 are incorporated into this assessment. Additional
details of the pertinent scientific literature published since the last review, as well as
selected earlier studies of particular interest, are included. This ISA thus serves to
update and revise the evaluation of the scientific evidence available at the time of the
completion of the 2006 O3 AQCD (U.S. EPA. 2006b). The current primary O3
standard includes an 8-hour (h) average (avg) standard set at 75 parts per billion
(ppb). The secondary standard for O3 is set equal to the primary standard. Further
information on the legislative and historical background for the O3 NAAQS is
contained in the Preface to this ISA.
This chapter summarizes and synthesizes the available scientific evidence and is
intended to provide a concise synopsis of the ISA conclusions and findings that best
inform consideration of the policy-relevant questions that frame this assessment
(U.S. EPA. 2009c). It includes:
• An integration of the evidence on the health effects associated with short- and
long-term exposure to O3, discussion of important uncertainties identified in
the interpretation of the scientific evidence, and an integration of health
evidence from the different scientific disciplines and exposure durations.
• An integration of the evidence on the welfare effects associated with exposure
to O3, including those associated with vegetation and ecosystems, and
discussion of important uncertainties identified in the interpretation of the
scientific evidence.
• Discussion of policy-relevant considerations, such as potentially at-risk
populations and concentration-response relationships and how they inform
selection of appropriate exposure metrics/indices.
2.1 ISA Development and Scope
EPA has a developed a robust, consistent, and transparent process for evaluating the
scientific evidence and drawing conclusions and causal judgments regarding air
pollution-related health and environmental effects. The ISA development process
includes literature search strategies, criteria for selecting and evaluating studies,
2-1
-------
approaches for evaluating the weight of the evidence, and a framework for making
causality determinations. The process and causality framework are described in more
detail in the Preamble to the ISA. This section provides a brief overview of the
process for development of this ISA.
EPA initiated the current review of the NAAQS for O3 on September 29, 2008, with
a call for information from the public (U.S. EPA, 2008g). Literature searches have
been conducted routinely since then to identify studies published since the last
review, focusing on studies published from 2005 (closing date for the previous
scientific assessment) through July 2011. References that were considered for
inclusion in this ISA can be found using the HERO website
(http://hero.epa.gov/ozone). This site contains HERO links to lists of references that
are cited in the ISA, as well as those that were considered for inclusion but not cited
in the ISA, with bibliographic information and abstracts.
This review has endeavored to evaluate all relevant data published since the last
review; this includes studies pertaining to the atmospheric science of O3, human
exposure to ambient O3, and health, ecological, climate and UV-B shielding effects
studies. These include studies that are related to concentration-response relationships,
mode(s) of action (MOA), and understanding of at-risk populations for effects of O3
exposure. Added to the body of research were EPA's analyses of air quality and
emissions data, and studies on atmospheric chemistry, transport, and fate of these
emissions using measurements and chemistry-transport models (CTMs).
Previous O3 AQCDs (U.S. EPA. 2006b. 1996a. b, 1984. 1978a) have included an
extensive body of evidence on both health and welfare effects of O3 exposure, as
well as an understanding of the atmospheric chemistry of O3 (U.S. EPA, 2006b).
In this ISA, the conclusions and key findings from previous reviews are summarized
at the beginning of each section, to provide the foundation for consideration of
evidence from recent studies. Results of key studies from previous reviews are
included in discussions or tables and figures, as appropriate, and conclusions are
drawn based on the synthesis of evidence from recent studies with the extensive
literature summarized in previous reviews.
The Preamble discusses the general framework for conducting the science
assessment and developing an ISA, including criteria for evaluating studies and
developing scientific conclusions. For selection of epidemiologic studies in the O3
ISA, particular emphasis is placed on those studies most relevant to the review of the
NAAQS. Studies conducted in the United States (U.S.) or Canada are discussed in
more detail than those from other geographical regions, and in regard to human
health, particular emphasis is placed on: (1) recent multi-city studies that employ
standardized analysis methods for evaluating effects of O3 and that provide overall
estimates for effects, based on combined analyses of information pooled across
multiple cities; (2) studies that help understand quantitative relationships between
exposure concentrations and effects; (3) new studies that provide evidence on effects
in at-risk populations; and (4) studies that consider and report O3 as a component of a
complex mixture of air pollutants. In evaluating toxicological and controlled human
exposure studies, emphasis is placed on studies using concentrations that are
2-2
-------
generally no greater than about an order of magnitude higher than ambient O3
concentrations currently found in many parts of the United States. Consideration of
studies important for evaluation of human exposure to ambient O3 places emphasis
on those evaluating the relationship between O3 measured at central site monitors
and personal exposure to ambient O3. Important factors affecting this relationship
include spatial and temporal variations in ambient O3 concentration, and time spent
outdoors, since penetrations of O3 into indoor environments may be limited.
This ISA uses a five-level hierarchy that classifies the weight of evidence for
causation:
• Causal relationship
• Likely to be a causal relationship
• Suggestive of a causal relationship
• Inadequate to infer a causal relationship, or
• Not likely to be a causal relationship
Beyond judgments regarding causality are questions relevant to quantifying health or
environmental risks based on the understanding of the quantitative relationships
between pollutant exposures and health or welfare effects. Once a determination is
made regarding the causal relationship between the pollutant and outcome category,
important questions regarding quantitative relationships include:
• What is the concentration-response, exposure-response, or dose-response
relationship?
• Under what exposure conditions (dose or concentration, duration, and pattern)
are effects observed?
• What populations or lifestages appear to be differentially affected, i.e., at
increased risk of O3-related health effects?
• What elements of the ecosystem (e.g., types, regions, taxonomic groups,
populations, functions, etc.) appear to be affected or are more sensitive to
effects?
This chapter summarizes and integrates the newly available scientific evidence that
best informs consideration of the policy-relevant questions that frame this
assessment. Section 2.2 discusses the trends in ambient concentrations and sources of
O3 and provides a brief summary of ambient air quality for past and current short-
and long-term exposure durations. Section 2.3 presents the evidence regarding
personal exposure to ambient O3 in outdoor and indoor microenvironments, and it
discusses the relationship between ambient O3 concentrations and personal exposure
to ambient O3. Section 2.4 provides a discussion of the dosimetry and modes of
action evidence for O3 exposure. Section 2.5 integrates the evidence for studies that
examine the health effects associated with short- and long-term exposure to O3 and
discusses important uncertainties identified in the interpretation of the scientific
evidence. A discussion of policy-relevant considerations, such as potentially at-risk
populations, and the O3 concentration-response relationship is also included in
2-3
-------
Section 2.5. Finally, Section 2.6 summarizes the evidence for welfare effects related
to O3 exposure, and Section 2.7 reviews the literature on the influence of
tropospheric O3 on climate and exposure to solar ultraviolet radiation.
2.2 Atmospheric Chemistry and Ambient Concentrations
2.2.1 Physical and Chemical Processes
Ozone in the troposphere is a secondary pollutant; it is formed by reactions of
precursor gases and is not directly emitted from specific sources. Ozone precursor
gases originate from both anthropogenic (i.e., man-made) and natural source
categories. Ozone attributed to anthropogenic sources is formed in the atmosphere by
photochemical reactions involving sunlight and precursor pollutants including
volatile organic compounds (VOCs), nitrogen oxides (NOX), and carbon monoxide
(CO). Ozone attributed to natural sources is formed through similar photochemical
reactions involving natural emissions of precursor pollutants from vegetation,
microbes, animals, biomass burning and lightning. An absolute distinction between
natural and anthropogenic sources of O3 precursors is often difficult to make in
practice, as human activities affect directly or indirectly emissions from what are
considered to be natural sources.
Ozone is present not only in polluted urban atmospheres but throughout the
troposphere, even in remote areas of the globe. The same basic processes involving
sunlight-driven reactions of NOX, VOCs and CO that occur in polluted urban air also
contribute to O3 formation throughout the troposphere1. In urban areas, NOX, VOCs,
and CO are all important precursors to O3 formation. In non-urban areas, biogenic
VOCs emitted from vegetation tend to be the most important precursors to O3
formation. In remote areas with little or no vegetation, and in general above the
planetary boundary layer (PEL, extending typically from 1 to 3 km above the
surface), methane—structurally the simplest VOC—and CO are the main carbon-
containing precursors to O3 formation. Ozone is subsequently removed from the
atmosphere through a number of gas phase reactions and deposition to surfaces.
Convective processes and turbulence transport O3 and other pollutants both upward
and downward throughout the PEL and the free troposphere above. If pollutants are
transported into the free troposphere above the PEL where winds are generally much
stronger than in the PEL, they can be transported over longer distances than they can
if they remained near the surface. Conversely, pollutants transported downward into
the PEL can add to pollution burdens there. The transport of pollutants downwind of
major urban centers is characterized by the development of urban plumes.
Meteorological conditions, small-scale circulation patterns induced by surface
characteristics and structures, localized chemistry, and topographic barriers can
1 These processes also lead to the formation of other photochemical products, such as peroxyacetyl nitrate, nitric acid, and sulfuric
acid, and to other compounds, such as formaldehyde and other carbonyl compounds.
2-4
-------
influence mixing on the intra-urban scale, resulting in variability in O3
concentrations across individual urban areas.
Apart from issues in understanding issues in O3 formation in areas such as the
Houston-Galveston-Brazoria airshed and the Northeast Corridor that are largely the
result of local pollution sources during summer, other issues have become apparent
in the past several years. Photochemically produced O3 concentrations that exceed
the level of the NAAQS have been observed in the Intermountain West in oil and gas
fields during specific meteorological conditions (i.e., fresh snow cover, low mixing
layer height trapping emissions) during winter. Because the mean tropospheric
lifetime of O3 is a few weeks, O3 can be transported from continent to continent.
Locations at high elevations are most susceptible to the intercontinental transport of
pollution, particularly during spring. Intrusions of stratospheric air containing high
O3 may also cause, or contribute significantly to, exceedances of levels of the
NAAQS for O3. These events occur mainly in the West during spring.
2.2.2 Background O3 Concentrations
In the context of a review of the NAAQS, it is useful to define background O3
concentrations in a way that distinguishes between concentrations that result from
precursor emissions that are relatively less controllable from those that are relatively
more controllable through U.S. policies. For this assessment, three definitions of
background O3 concentrations are considered, including (1) United States (U.S.)
background (simulated O3 concentrations that would exist in the absence of
anthropogenic emissions from the U.S.), (2) North American (NA) background
(simulated O3 concentrations that would exist in the absence of anthropogenic
emissions from the U.S., Canada and Mexico), and (3) natural background
(simulated O3 concentrations in the absence of all anthropogenic emissions globally).
Each definition of background O3 includes contributions resulting from emissions
from natural sources (e.g., stratospheric intrusions, wildfires, biogenic methane and
more short-lived VOC emissions) throughout the globe. Differences among these
definitions reflect differences in the inclusion of geographic regions that are sources
of anthropogenic precursors. These definitions are used to inform policy
considerations regarding the current or potential alternative standards. Note also there
is no chemical difference between background O3, and O3 attributable to U.S. or
North American anthropogenic sources.
It is important to note that since background O3 concentrations as defined above are
a hypothetical construct that cannot be directly measured, the range of background
O3 concentrations must be estimated using CTMs. This is because observations
within the U.S.—even at relatively remote monitoring sites—can be impacted by
transport from anthropogenic sources within the U.S. or within the rest of North
America (if a North American background is adopted).
Figure 2-1 shows spring and summer mean, maximum daily 8-h average O3
concentrations simulated by the GEOS-Chem global CTM for the base case (top
2-5
-------
panel), which includes all sources of O3 precursors and can be compared to
observations; the U.S. background (middle panel) and the North American (NA)
background (bottom panel). Seasonal mean values for the entire continental United
States are shown above each panel. In general, for most areas of the United States
simulated seasonal mean 8-h max O3 concentrations in the base case (top panel) are
within a fewppb of those observed at rural Clean Air Status and Trends Network
(CASTNET) monitoring sites, for which details are given in the next section.
In moving from the top to the bottom panel, it can be seen that calculated
concentrations decrease. Note also that mean base case concentrations (including
U.S., Canadian and Mexican sources) are higher in the summer than in the spring,
but that mean U.S. background and NA background concentrations are higher in
spring than in summer, reflecting the increased importance of sources such as
intercontinental transport and stratospheric intrusions.
As can be seen from the middle and bottom panels in Figure 2-1, estimated
U.S. background and NA background concentrations tend to be higher in the West
(particularly in the Intermountain West) and in the Southwest compared to the East
in both spring and summer. U.S. background and NA background concentrations
tend to be highest in the Southwest during summer in the GEOS-Chem model, and
are driven in large part by lightning NOX. As can be seen from Figure 2-1 (middle
panel), highest U.S. background concentrations (in the U.S.) are found over the
Northern Tier of New York State. High values are also found in other areas
bordering Canada and Mexico. NA background concentrations (bottom panel) are on
average ~3 ppb higher than U.S. background concentrations (middle panel) during
spring and summer across the United States. For March through August 2006, mean
NA background O3 concentrations of 29 ± 8 ppb at low elevation (<1,500 meters)
and 40 ± 8 ppb at high elevation (>1,500 meters) were calculated. Corresponding
natural background concentrations (not shown) range from 18 ± 6 ppb to 27 ± 6 ppb.
It should also be noted that methane is an important contributor to NA background
O3, accounting for slightly less than half of the increase in background since the pre-
industrial era and whose relative contribution is projected to grow in the future.
Note that the calculations of background concentrations presented in Chapter 1 were
formulated to answer the question, "what would O3 concentrations be if there were
no anthropogenic sources." This is different from asking, "how much of the O3
measured or simulated in a given area is due to background contributions." Because
of potentially strong non-linearities—particularly in many urban areas—these
estimates should not be used by themselves to answer the second question posed
above. The extent of these non-linearities will generally depend on location and time,
the strength of concentrated sources, and the nature of the chemical regime.
2-6
-------
Base: Spring (49 ppbv)
111 Hid
Longitude (degrees)
Base: Summer (52 ppbv)
1211 lilt HMI »\ Ml -7)1
Longitude (degrees)
USB: Spring (36 ppbv)
USB: Summer (33 ppbv)
— 130 -1211 -1111 -IIHJ -!KI -Ml -711
Longitude (degrees)
-Iffl -120 -110 -KM) -30 -SO -TO
Longitude (degrees)
NAB: Spring (33 ppbv)
NAB: Summer (30 ppbv)
— 1311 -1211 -Iln -IIHI -'«) -Mil -71)
Longitude (degrees)
-1.10 -12(1 -111! -1IHJ -HIP -S(l -711
Longitude (degrees)
15
25
35
45
55
65
Note: Seasonal mean daily maximum 8-h avg O3 concentrations were calculated by GEOS-Chem for the base case (top, Base),
United States background (middle, USB) and North American Background (bottom, NAB).
Values in parentheses (above each map) refer to continental U.S. means, and are shown in the color bar as black squares for
summer and white squares for spring.
Source: Adapted from Zhang et al. (2011).
Figure 2-1 Mean daily average maximum 8-h avg Os concentrations in surface
air, for spring and summer 2006.
2-7
-------
2.2.3 Monitoring
The federal reference method (FRM) for O3 measurement is based on the detection
of chemiluminescence resulting from the reaction of O3 with ethylene gas. However,
almost all of the state and local air monitoring stations (SLAMS) that reported data to
the EPA's Air Quality System (AQS) database from 2005 to 2009 used the federal
equivalence method (FEM) UV absorption photometer. Relative to FRMs, FEMs
must satisfy precision and bias requirements to be accepted as alternative methods
for sampling and analyzing ambient air. More than 96% of O3 monitors met these
precision and bias requirements for designation as an FEM during this period.
In 2010, there were 1,250 SLAMS O3 monitors reporting data to AQS. Ozone
monitoring is required at SLAMS sites during the local "ozone season" which varies
by state. In addition, National Core (NCore) is a new multipollutant monitoring
network implemented to meet multiple monitoring objectives and each state is
required to operate at least one NCore site. The NCore network consists of 60 urban
and 20 rural sites nationwide (see Figure 3-21 and Figure 3-22). The densest
concentrations of O3 sites are located in California and the eastern half of the U.S.
The Clean Air Status and Trends Network (CASTNET) is a regional monitoring
network established to assess trends in acidic deposition and also provides
concentration measurements of O3. CASTNET O3 monitors operate year round and
are primarily located in rural areas; in 2010, there were 80 CASTNET sites reporting
O3 data to AQS. The National Park Service (NFS) operates 23 CASTNET sites in
national parks and other Class-i areas, and provided data to AQS from 20 additional
Portable Ozone Monitoring Systems (POMS) in 2010 (see Figure 3-22). Compared
to urban-focused monitors, rural-focused monitors are relatively scarce across the
U.S.
2.2.4 Ambient Concentrations
Ozone is the only photochemical oxidant other than NO2 that is routinely monitored
and for which a comprehensive database exists. Other photochemical oxidants are
typically only measured during special field studies. The concentration analyses in
Chapter 3_ are limited to widely available O3 data obtained directly from AQS for the
period from 2007 to 2009. The median 24-h average, 8-h daily max, and 1-h daily
max O3 concentrations across all U.S. sites reporting data to AQS between 2007 and
2009 were 29, 40, and 44 ppb, respectively.
To investigate O3 variability in urban areas across the U.S., 20 combined statistical
areas (CSAs) were selected for closer analysis based on their importance in O3
epidemiology studies and on their location. Several CSAs had relatively little spatial
variability in daily maximum 8-h avg O3 concentrations (e.g., inter-monitor
correlations ranging from 0.61 to 0.96 in the Atlanta, GA, CSA) while other CSAs
exhibited considerably more variability in O3 concentrations (e.g., inter-monitor
correlations ranging from -0.06 to 0.97 in the Los Angeles, CA, CSA). Uncertainties
2-8
-------
resulting from the spatial variability in O3 concentration fields should be considered
when using data from the network of ambient O3 monitors to approximate
community-scale exposures, since community exposure may not be well-represented
when monitors cover large areas with multiple subcommunities having different
sources and topographies. However, studies evaluating the influence of monitor
selection on epidemiologic study results have found that O3 effect estimates are
similar across different spatial scales and monitoring sites (Section 4.6.2.1).
To investigate O3 variability in rural settings across the U.S., six focus areas were
selected for closer analysis based on the impact of O3 or O3 precursor transport from
upwind urban areas. The selected rural focus area with the largest number of
available AQS monitors was Great Smoky Mountains National Park where the May-
September median 8-h daily max O3 concentration ranged from 47 ppb at the lowest
elevation (564 meters) site to 60 ppb at the highest elevation (2,021 meters) site.
Correlations between sites within each rural focus area ranged from 0.78 to 0.92.
Ozone in rural areas is produced from emissions of O3 precursors emitted directly
within the rural areas, from emissions in urban areas that react and chemically
transform during transport, and from occasional stratospheric intrusions. Factors
contributing to variations observed within these rural focus areas include proximity
to local O3 precursor emissions, local scale circulations related to topography, and
possibly stratospheric intrusions as a function of elevation. In addition, O3 tends to
persist longer in rural than in urban areas as a result of less chemical scavenging
from other primary pollutants. This results in a more uniform O3 concentration
throughout the day and night without the typical nocturnal decrease in O3
concentration observed in urban areas. Persistently high O3 concentrations observed
at many of the rural sites investigated here indicate that cumulative exposures for
humans and vegetation in rural areas can frequently exceed cumulative exposures in
urban areas.
Nation-wide surface-level O3 concentrations have declined over the last decade, with
a particularly noticeable decrease between 2003 and 2004 coinciding with NOX
emissions reductions resulting from implementation of the NOX State
Implementation Plan (SIP) Call rule, which began in 2003 and was fully
implemented in 2004. This rule was designed to reduce NOX emissions from power
plants and other large combustion sources in the eastern United States. The largest
density of individual monitors showing downward trends in O3 concentrations over
the last decade occur in the Northeast where this rule was focused. In addition to a
downward trend, the nation-wide surface-level O3 concentration data also show a
general tightening of the distribution across sites. In contrast to the majority of U.S.
surface-level monitors reporting downward trends, a few surface-level monitors and
elevated observations along the Pacific Coast have shown increases in O3
concentrations in recent years, possibly resulting from intercontinental transport from
Asia. As noted in the 2006 O3 AQCD (U.S. EPA, 2006b), trends in national parks
and rural areas are similar to nearby urban areas, reflecting the regional nature of O3
pollution.
2-9
-------
Since O3 is a secondary pollutant, it is not expected to be highly correlated with
primary pollutants such as CO and NOX. Furthermore, O3 formation is strongly
influenced by meteorology, entrainment, and transport of both O3 and O3 precursors,
resulting in a broad range in correlations with other pollutants which can vary
substantially with season. Temporal relationships between 8-h daily max O3 and
other criteria pollutants exhibit mostly negative correlations in the winter and mostly
positive correlations in the summer. As a result, statistical analyses that may be
sensitive to correlations between copollutants need to take seasonality into
consideration, especially when O3 is being investigated.
2.3 Human Exposure
The widespread presence of O3 in the environment results in exposure as people
participate in normal daily activities. Personal exposure measurements have been
found to be moderately associated with fixed-site ambient O3 concentrations,
although a number of factors affect the relationship between ambient concentration
and personal exposure. These include: infiltration of ambient O3 into indoor
microenvironments, which is driven by air exchange rate; time spent outdoors and
activity pattern, which includes changes in personal behavior by some populations to
avoid exposure to O3, and influences of lifestage; and the variation in O3
concentrations at various spatial and temporal scales. Personal exposure to O3 is
moderately correlated with ambient O3 concentration, as indicated by studies
reporting correlations generally in the range of 0.3-0.8 (Table 4-2). This suggests that
ambient monitor concentrations are representative of day-to-day changes in personal
exposure to ambient O3. Some studies report lower personal-ambient correlations, a
result attributable in part to low building air exchange rates and O3 concentrations
below the personal sampler detection limit. Low correlations can also occur for
individuals or populations spending increased time indoors. In contrast to correlation,
which represents the temporal association between exposure and concentration, the
magnitude of exposure can be represented as the ratio between personal exposure and
ambient concentration. This ratio varies widely depending on activity patterns,
housing characteristics, and season. Personal-ambient ratios are typically 0.1-0.3 for
sampling durations of several hours to several days, although individuals spending
substantial time outdoors (e.g., outdoor workers) have shown much higher ratios
(0.5-0.9) (Table 4-3). Since there are relatively few indoor sources of O3, and
because of reactions of O3 with indoor surfaces and airborne constituents, indoor O3
concentrations are often substantially lower than outdoor concentrations
(Section 4.3.2). The lack of indoor sources also suggests that fluctuations in ambient
O3 may be primarily responsible for changes in personal exposure, even under low-
ventilation, low-concentration conditions.
Another factor that can influence the pattern of exposure is the tendency for people to
avoid O3 exposure by altering their behavior (e.g., reducing outdoor activity levels or
time spent being active outdoors) on high-O3 days. Activity pattern has a substantial
effect on ambient O3 exposure, with time spent outdoors contributing to increased
2-10
-------
exposure (Section 4.4.2). Air quality alerts and public health recommendations
induce reductions in time spent outdoors on high-O3 days among some lifestages and
populations, particularly for children, older adults, and people with respiratory
problems. Such effects are less pronounced in the general population. Limited
evidence from an epidemiologic study conducted in the 1990s in Los Angeles, CA
reports increased asthma hospital admissions among children and older adults when
O3 alert days (1-h max O3 concentration >200 ppb) were excluded from the analysis
of daily hospital admissions and O3 concentrations (presumably thereby eliminating
averting behavior based on high O3 forecasts). The lower rate of admissions
observed when alert days were included in the analysis suggests that estimates of
health effects based on concentration-response functions that do not account for
averting behavior may be biased toward the null.
Variations in O3 concentrations occur over multiple spatial and temporal scales. Near
roadways, O3 concentrations are reduced due to reaction with NO and other species
(Section 4.3.4.2). Over spatial scales of a few kilometers and away from roads, O3
can be somewhat more homogeneous due to its formation as a secondary pollutant,
while over scales of tens of kilometers, additional atmospheric processing can result
in higher concentrations downwind of an urban area. Although local-scale variability
impacts the magnitude of O3 concentrations, O3 formation rates are influenced by
factors that vary over larger spatial scales, such as temperature (Section 3.2),
suggesting that urban monitors may track one another temporally, but miss small-
scale variability. This variation in concentrations changes the pattern of exposure
people experience as they move through different microenvironments and affects the
magnitude of exposures in different locations within an urban area. The various
factors affecting exposure patterns and quantification of exposure result in
uncertainty which can contribute to exposure measurement error in epidemiologic
studies, which typically use fixed-site monitor data as an indicator of exposure. Low
personal-ambient ratios result in attenuation of the magnitude of the exposure-based
effect estimate or response function relative to the concentration-based response
function, although the statistical association is similar for concentration- and
exposure-based effect estimates if the ratio is approximately constant over time. Low
personal-ambient correlations are a source of exposure error for epidemiologic
studies, tending to obscure the presence of potential thresholds, bias effect estimates
toward the null, and widen confidence intervals, and this impact may be more
pronounced among populations spending substantial time indoors. The impact of this
exposure error may tend more toward widening confidence intervals than biasing
effect estimates, since epidemiologic studies evaluating the influence of monitor
selection indicate that effect estimates are similar across different spatial averaging
scales and monitoring sites. In addition, in examinations of respiratory endpoints in
epidemiologic studies, associations were similar in magnitude across analyses using
several different concentration metrics to estimate exposure, such as on-site
measurements, closest-site measurements, and multi-site averages. These exposure
estimation methods likely vary in how well ambient O3 concentrations represent
personal exposures and between-subject variability in exposures. Respiratory effects
were observed with ambient O3 concentrations found to have stronger personal-
ambient relationships, including those measured on-site during long periods of
2-11
-------
outdoor activity. However, such effects were also found with ambient O3
measurements expected to have weaker personal-ambient relationships, including
those measured at home or school, measured at the closest site, averaged from
multiple community sites, and measured at a single site. Overall, there was no clear
indication that a particular method of exposure estimation produced stronger
findings.
2.4 Dosimetry and Mode of Action
Upon inspiration, O3 uptake in the respiratory tract is affected by a number of factors
including respiratory tract morphology, and breathing route, frequency, and volume.
Additionally, physicochemical properties of O3 itself and how it is transported, as
well as the physical and chemical properties of the extracellular lining fluid (ELF)
and tissue layers in the respiratory tract can influence O3 uptake. Experimental
studies and models have suggested that there are differences between the total
absorption of O3 from the inhaled air and the O3 dose reaching the respiratory tract
tissues. The total O3 absorption gradually decreases with distal progression into the
respiratory tract allowing for a proportionally large concentration of O3 to be
absorbed in the upper respiratory tract; thus, causing the nasal membranes to be a
potential target site of O3-induced injury. In contrast, the primary site of O3 delivery
to the lung epithelium is believed to be the centriacinar region or the junction of the
conducting airways with the gas exchange region. In addition, as a large
concentration of the O3 absorbed by the respiratory tract is absorbed in the upper
respiratory tract, the nasal membranes are another potential site of O3-induced injury.
Ozone uptake is sensitive to a number of factors including tidal volume, breathing
frequency, O3 concentration, and exposure time. Interindividual variability also
accounts for a large amount of the variability in local dose due to differences in
pulmonary physiology, anatomy, and biochemistry. An increase in tidal volume and
breathing frequency are both associated with increased physical activity. These
changes and a switch to oronasal breathing during exercise result in deeper
penetration of O3 into the lower respiratory tract in part due to less oral versus nasal
uptake efficiency. For these reasons, increased physical activity acts to move the
maximum tissue dose of O3 distally in the respiratory tract and more into the alveolar
region.
The ELF is a complex mixture of lipids, proteins, and antioxidants that serves as the
first barrier and target for inhaled O3 (see Figure 5-7). Distinct products with diverse
reactivity (i.e., secondary oxidation products), are mainly formed by reactions of O3
with soluble ELF components. The thickness of the ELF is an important determinant
of the dose of O3 to the tissues; a thicker ELF generally results in a lower dose of O3
to the tissues. Additionally, the quenching ability and the concentrations of
antioxidants and other ELF components are determinants of the formation of
secondary oxidation products. These reactions appear to limit interaction of O3 with
underlying tissues and to reduce penetration of O3 distally into the respiratory tract.
2-12
-------
In addition to contributing to the driving force for O3 uptake, formation of secondary
oxidation products contributes to oxidative stress which can lead to cellular injury
and altered cell signaling in the respiratory tract. Secondary oxidation products
initiate pathways (see Figure 5-8) that provide the mechanistic basis for short- and
long-term health effects described in detail in Chapters 6 and 7. Other key events
involved in the mode of action of O3 in the respiratory tract include the activation of
neural reflexes, initiation of inflammation, alterations of epithelial barrier function,
sensitization of bronchial smooth muscle, modification of innate and adaptive
immunity, and airways remodeling. Another key event, systemic inflammation and
vascular oxidative/nitrosative stress, may be critical to the extrapulmonary effects of
03.
Secondary oxidation products can transmit signals to respiratory tract cells resulting
in the activation of neural reflexes. Nociceptive sensory nerves mediate the
involuntary truncation of respiration, resulting in decreases in lung function
(i.e., FVC, FEVi, and tidal volume), and pain upon deep inspiration. Studies
implicate TRPA1 receptors on bronchial C-fibers in this reflex. Another neural reflex
involves vagal sensory nerves, which mediate a mild increase in airways obstruction
(i.e., bronchoconstriction) following exposure to O3 via parasympathetic pathways.
Substance P release from bronchial C-fibers and the SP-NK receptor pathway can
also contribute to this response.
Secondary oxidation products also initiate the inflammatory cascade following
exposure to O3. Studies have implicated eicosanoids, chemokines and cytokines,
vascular endothelial adhesion molecules, and tachykinins in mediating this response.
Inflammation is characterized by airways neutrophilia as well as the influx of other
inflammatory cell types. Recent studies demonstrate a later phase of inflammation
characterized by increased numbers of macrophages, which is mediated by
hyaluronan. Inflammation further contributes to O3-induced oxi dative stress.
Alteration of the epithelial barrier function of the respiratory tract also occurs as a
result of O3-induced secondary oxidation product formation. Increased epithelial
permeability may lead to enhanced sensitization of bronchial smooth muscle,
resulting in airways hyperresponsiveness (AHR). Neurally-mediated sensitization
also occurs and is mediated by cholinergic postganglionic pathways and bronchial C-
fiber release of substance P. Recent studies implicate hyaluronan and Toll-like
receptor 4 (TLR4) signaling in bronchial smooth muscle sensitization, while earlier
studies demonstrate roles for eicosanoids, cytokines, and chemokines.
Evidence is accumulating that exposure to O3 modifies innate and adaptive immunity
through effects on macrophages, monocytes, and dendritic cells. Enhanced antigen
presentation, adjuvant activity, and altered responses to endotoxin have been
demonstrated. TLR4 signaling contributes to some of these responses. Effects on
innate and adaptive immunity can result in both short- and longer-term consequences
related to the exacerbation and/or induction of asthma and to alterations in host
defense.
2-13
-------
Airways remodeling has been demonstrated following chronic and/or intermittent
exposure to O3 by mechanisms that are not well understood. However, the TGF-(3
signaling pathway has recently been implicated in O3-induced deposition of collagen
in the airways wall. These studies were conducted in adult animal models and their
relevance to effects in humans is unknown.
Evidence is also accumulating that O3 exposure results in systemic inflammation and
vascular oxidative/nitrosative stress. The release of diffusible mediators from the
O3-exposed lung into the circulation can initiate or propagate inflammatory
responses in the vascular or in systemic compartments. This may provide a
mechanistic basis for extrapulmonary effects, such as vascular dysfunction.
Both dosimetric and mechanistic factors contribute to the understanding of
inter-individual variability in response. Inter-individual variability is influenced by
variability in respiratory tract volume and thus surface area, breathing route, certain
genetic polymorphisms, pre-existing conditions and disease, nutritional status,
lifestages, attenuation, and co-exposures. In particular, very young individuals may
be sensitive to developmental effects of O3 since studies in animal models
demonstrated altered development of lung and immune system.
Some of these factors are also influential in understanding species homology and
sensitivity. Qualitatively, animal models exhibit a similar pattern of tissue dose
distribution for O3 with the largest tissue dose delivered to the centriacinar region.
However, due to anatomical and biochemical respiratory tract differences, the actual
O3 dose delivered differs between humans and animal models. Animal data obtained
in resting conditions underestimates the dose to the respiratory tract tissue relative to
exercising humans. Further, it should be noted that, with the exception of airways
remodeling, the mechanistic pathways discussed above have been demonstrated in
both animals and human subjects in response to the inhalation of O3. Even though
interspecies differences limit quantitative comparison between species, the short- and
long-term functional responses of laboratory animals to O3 appear qualitatively
homologous to those of the human making them a useful tool in determining
mechanistic and cause-effect relationships with O3 exposure. Furthermore, animal
studies add to a better understanding of the full range of potential O3-mediated
effects.
2.5 Integration of Ozone Health Effects
This section evaluates the evidence from toxicological, controlled human exposure,
and epidemiologic studies (which examined the health effects associated with short-
and long-term exposure to O3,) and summarizes the main conclusions of this
assessment regarding the health effects of O3 and the concentrations at which those
effects are observed. The results from the health studies, supported by the synthesis
of atmospheric chemistry (see Section 2.2) and exposure assessment (see
Section 2.3) studies, contribute to the causal determinations made for the health
2-14
-------
outcomes discussed in this assessment (see Preamble to this document for details on
the causal framework).
Epidemiologic studies generally present O3-related effect estimates for mortality and
morbidity health outcomes based on an incremental change in exposure, traditionally
equal to the interquartile range in O3 concentrations or some other arbitrary value
(e.g., 10 ppb). Additionally, various averaging times are used in O3 epidemiologic
studies, with the three most common being the maximum 1-hour average within a
24-hour period (1-h max), the maximum 8-hour average within a 24-hour period
(8-h max), and 24-hour average (24-h avg). For the purpose of presenting results
from studies that use different exposure metrics, EPA consistently applies the same
O3 increments to facilitate comparisons between the results of various studies that
may use different indices. These increments were derived using the nationwide
distributional data for O3 monitors in U.S. Metropolitan Statistical Areas. They are
representative of a low-to-high change in O3 concentrations and were approximated
on the basis of annual mean to 95th percentile differences (Langstaff, 2003).
Therefore, throughout Chapter 6, efforts were made to standardize O3-related effect
estimates using the increments of 20 ppb for 24-h avg, 30 ppb for 8-h max, and
40 ppb for 1-h max O3 concentrations, except as noted. In long-term exposure
studies, typically, O3 concentrations are lower and less variable when averaged
across longer exposure periods, and differences due to the use of varying averaging
times (e.g., 1-h max, 24-h avg) become less apparent. As such, in the long-term
exposure chapter (Chapter 7) an increment of 10 ppb was consistently applied across
studies, regardless of averaging time, to facilitate comparisons between the results
from these studies.
2.5.1 Conclusions from Previous Os AQCDs
The 2006 O3 AQCD concluded that there was clear, consistent evidence of a causal
relationship between short-term O3 exposure and respiratory health effects (U.S.
EPA, 2006b). The causal relationship for respiratory health effects was substantiated
by the coherence of effects observed across controlled human exposure,
epidemiologic, and toxicological studies indicating effects of short-term O3
exposures on a range of respiratory health endpoints from respiratory tract
inflammation to respiratory-related emergency department (ED) visits and hospital
admissions.
Across disciplines, short-term O3 exposures induced or were associated with
statistically significant declines in lung function. An equally strong body of evidence
from controlled human exposure and toxicological studies demonstrated O3-induced
inflammatory responses, increased epithelial permeability, and airway
hyperresponsiveness (both specific and nonspecific). Toxicological studies provided
additional evidence for O3-induced impairment of host defenses. Combined, these
findings from experimental studies provided support for epidemiologic evidence, in
which short-term increases in ambient O3 concentration were consistently associated
2-15
-------
with increases in respiratory symptoms and asthma medication use in children with
asthma, respiratory-related hospital admissions, and asthma-related ED visits.
Although O3 was consistently associated with nonaccidental and cardiopulmonary
mortality, the contribution of respiratory causes to these findings was uncertain.
Collectively, there is a vast amount of evidence spanning several decades that
demonstrated that exposure to O3 induces a range of respiratory effects. The majority
of this evidence was derived from studies investigating short-term exposure
(i.e., hours to weeks) to O3. The combined evidence across disciplines led to the
causal relationship between short-term O3 exposure and respiratory effects reported
in the 2006 O3 AQCD.
Mechanistic evidence for the effects of O3 on the respiratory system was
characterized in the 1996 O3 AQCD (U.S. EPA. 1996aX which identified O3-induced
changes in a variety of lung lipid species whose numerous biologically active
metabolites, in turn, can affect host defenses, lung function, and the immune system.
As summarized in Section 2.4 and fully characterized in Chapter 5, key events in the
toxicity pathway of O3 have been identified in humans and animal models. They
include formation of secondary oxidation products, activation of neural reflexes,
initiation of inflammation, alteration of epithelial barrier function, sensitization of
bronchial smooth muscle, modification of innate/adaptive immunity, airways
remodeling, and systemic inflammation and oxidative/nitrosative stress.
2.5.2 Summary of Causal Determinations
Recent studies support or build upon the strong body of evidence presented in the
1996 and 2006 O3 AQCDs that short-term O3 exposure is causally associated with
respiratory health effects. Recent controlled human exposure studies demonstrate
statistically significant group mean decreases in pulmonary function to exposures as
low as 60-70 ppb O3 in young, healthy adults, and are supported by the strong,
cumulative evidence from epidemiologic studies. Equally strong evidence
demonstrated associations of ambient O3 with respiratory hospital admissions and
ED visits across the U.S., Europe, and Canada. Most effect estimates ranged from a
1.6 to 5.4% increase in daily respiratory-related ED visits or hospital admissions in
all-year analyses for unit increases1 in ambient O3 concentrations. Several multicity
studies and a multicontinent study reported associations between short-term increases
in ambient O3 concentrations and increases in respiratory mortality. This evidence is
supported by a large body of individual-level epidemiologic panel studies that
demonstrate associations of ambient O3 with respiratory symptoms in children with
asthma. Further support is provided by recent studies that found O3-associated
increases in indicators of airway inflammation and oxidative stress in children with
asthma. Across respiratory endpoints, evidence indicates antioxidant capacity may
modify the risk of respiratory morbidity associated with O3 exposure. The potentially
elevated risk of populations with diminished antioxidant capacity and the reduced
1 Effect estimates were standardized to a 40-, 30-, and 20-ppb unit increase for 1-h max, 8-h max, and 24-h avg O3.
2-16
-------
risk of populations with enhanced antioxidant capacity identified in epidemiologic
studies is strongly supported by similar findings from controlled human exposure
studies and by evidence that characterizes O3-induced decreases in intracellular
antioxidant levels as a mode of action for downstream effects. By demonstrating
O3-induced airway hyperresponsiveness, decreased pulmonary function, allergic
responses, lung injury, impaired host defense, and airway inflammation,
toxicological studies have characterized O3 modes of action and provided biological
plausibility for epidemiologic associations of ambient O3 concentrations with lung
function and respiratory symptoms, hospital admissions, ED visits, and mortality.
Together, the evidence integrated across controlled human exposure, epidemiologic,
and toxicological studies and across the spectrum of respiratory health endpoints
continues to demonstrate that there is a causal relationship between short-term
O3 exposure and respiratory health effects.
The epidemiologic evidence for a relationship between long-term O3 exposure and
respiratory health effects (including respiratory symptoms, new-onset asthma, and
respiratory mortality) is contributed by recent studies that evaluate the associations
between long-term exposure to O3 and respiratory effects that demonstrate
interactions between exercise or different genetic variants and both new-onset asthma
in children and increased respiratory symptom effects in individuals with asthma.
While the evidence is limited, a U.S. multicommunity prospective cohort
demonstrates that asthma risk is affected by interactions among genetic variability,
environmental O3 exposure, and behavior. The evidence relating new-onset asthma
to long-term O3 exposure is supported by toxicological studies of asthma in
monkeys. This nonhuman primate evidence of O3-induced changes supports the
biologic plausibility of long-term exposure to O3 contributing to the effects of
asthma in children. Early life O3 exposure can alter airway development and lead to
the development of asthma. Other recent epidemiologic studies provide coherent
evidence for long-term O3 exposure and respiratory effects such as first asthma
hospitalization, respiratory symptoms in asthmatics, and respiratory mortality.
Generally, the epidemiologic and toxicological evidence provides a compelling case
that supports the hypothesis that a relationship exists between long-term exposure to
ambient O3 and measures of respiratory health effects and mortality. The evidence
for short-term exposure to O3 and effects on respiratory endpoints provides
coherence and biological plausibility for the effects of long-term exposure to O3.
Building upon that evidence, the more recent epidemiologic evidence, combined with
toxicological studies in rodents and nonhuman primates, provides biologically
plausible evidence that there is likely to be a causal relationship between long-
term exposure to O3 and respiratory health effects.
In past O3 AQCDs the effects of short- term exposure to O3 on the cardiovascular
system could not be thoroughly evaluated due to the paucity of information available.
However, studies investigating O3-induced cardiovascular events have advanced in
the last two decades. Animal toxicological studies, although limited in number,
demonstrate O3-induced cardiovascular effects; specifically enhanced
ischemia/reperfusion (I/R) injury, disrupted NO-induced vascular reactivity,
decreased cardiac function, and increased heart rate variability (HRV). These effects
2-17
-------
are consistent with cardiovascular system effects observed after long-term O3
exposure, such as increased vascular disease. These effects may, in part, correspond
to the alteration of the autonomic nervous system or to the development and
maintenance of systemic oxidative stress and a proinflammatory environment that
can result from pulmonary inflammation. Controlled human exposure studies provide
some coherence with the evidence from animal toxicological studies, by
demonstrating increases and decreases in HRV following relatively low (120 ppb
during rest) and high (300 ppb with exercise) O3 exposures, respectively. Controlled
human exposure studies also support the animal toxicology studies by demonstrating
Os-induced effects on blood biomarkers of systemic inflammation and oxi dative
stress as well as changes in biomarkers suggestive of a pro-thrombogenic response to
O3. The experimental evidence provides initial biological plausibility for the
consistently positive associations observed across multiple epidemiologic studies of
short-term O3 exposure and cardiovascular mortality. However, epidemiologic
studies generally do not observe associations between short-term exposure to O3 and
cardiovascular morbidity; studies of cardiovascular-related hospital admissions and
ED visits and other various cardiovascular effects did not find consistent evidence of
a relationship with O3 exposure. The lack of coherence between the results from
studies that examined associations between short-term O3 exposure and
cardiovascular morbidity and subsequently cardiovascular mortality complicate the
interpretation of the overall evidence for O3-induced cardiovascular effects. Overall,
animal toxicological studies demonstrate O3-induced cardiovascular effects, and
support to the strong body of evidence indicating O3-induced cardiovascular
mortality. Animal toxicological and controlled human exposure studies provide
evidence for biologically plausible mechanisms underlying these O3-induced
cardiovascular effects. However, a lack of coherence with epidemiologic studies of
cardiovascular morbidity remains an important uncertainty. Taken together, the
overall body of evidence across disciplines indicates that there is likely to be a
causal relationship between short-term exposures to O3 and cardiovascular
effects.
The 2006 O3 AQCD concluded that the overall body of evidence was highly
suggestive that short-term exposure to O3 directly or indirectly contributes to
nonaccidental and cardiopulmonary-related mortality, but additional research was
needed to more fully establish underlying mechanisms by which such effects occur.
The evaluation of recent multicity studies and a multicontinent study that examined
the association between short-term increases in ambient O3 concentration and
mortality found evidence that supports the conclusions of the 2006 O3 AQCD. These
recent studies reported consistent positive associations between short-term increases
in ambient O3 concentration and total (nonaccidental) mortality, with associations
being stronger during the warm season, as well as provided additional support for
associations between O3 concentrations and cardiovascular mortality being similar or
larger in magnitude compared to respiratory mortality. Additionally, these new
studies examined previously identified areas of uncertainty in the O3-mortality
relationship, and provide additional evidence supporting an association between
short-term O3 exposure and mortality. Taken together, the body of evidence indicates
2-18
-------
that there is likely to be a causal relationship between short-term O3 exposures
and total mortality.
The 2006 O3 AQCD concluded that an insufficient amount of evidence existed to
suggest a causal relationship between long-term O3 exposure and mortality (U.S.
EPA, 2006b). A synthesis of the recent and earlier evidence reveals that the strongest
evidence for an association between long-term exposure to ambient O3
concentrations and mortality is derived from associations for respiratory mortality
that remained robust after adjusting for PM2.5 concentrations. There is inconsistent
evidence for an association between long-term exposure to ambient O3 and
cardiopulomary mortality, with several analyses from the American Cancer Society
(ACS) cohort reporting some positive associations, while other studies reported no
association. There is generally limited evidence for an association with long-term
exposure to ambient O3 and total mortality. The findings for respiratory mortality are
consistent and coherent with the evidence from epidemiologic, controlled human
exposure, and animal toxicological studies for the effects of short- and long-term
exposure to O3 on respiratory effects. Respiratory mortality is a relatively small
portion of total mortality [about 7.6% of all deaths in 2010 were due to respiratory
causes (Murphy et al.. 2012)1. thus it is not surprising that the respiratory mortality
signal may be difficult to detect in studies of cardiopulmonary or total mortality.
Based on the recent evidence for respiratory mortality along with limited evidence
for total and cardiopulmonary mortality, the evidence is suggestive of a causal
relationship between long-term O3 exposures and total mortality.
In past O3 AQCDs the effects of long-term exposure to O3 on the cardiovascular
system could not be thoroughly evaluated due to the paucity of information available.
However, studies investigating O3-induced cardiovascular events have advanced in
the last two decades. Animal toxicological studies provide evidence for long-term O3
exposure leading to cardiovascular morbidity, including increased vascular disease.
There is limited, inconsistent evidence for cardiovascular morbidity in epidemiologic
studies examining long-term exposure to O3. Overall, animal toxicological studies
provide some evidence for O3-induced cardiovascular effects, but the effects
observed were not consistently supported by controlled human exposure studies or
epidemiologic studies. Thus, the overall body of evidence across disciplines is
suggestive of a causal relationship between long-term exposures to O3 and
cardiovascular effects.
In the 2006 O3 AQCD, there were a number of health effects for which an
insufficient amount of evidence existed to adequately characterize the relationships
with exposure to O3. However, recent evidence suggests that O3 may impart health
effects through exposure durations and biological mechanisms not previously
considered. For example, recent toxicological studies add to earlier evidence that
short- and long-term exposures to O3 can produce a range of effects on the central
nervous system and behavior. Additionally, an epidemiologic study demonstrated
that long-term exposure to O3 affects memory in humans as well. Together the
evidence from studies of short- and long-term exposure to O3 is suggestive of a
causal relationship between O3 exposure and central nervous system effects.
There is also limited though positive toxicological evidence for O3-induced
2-19
-------
developmental effects. Limited epidemiologic evidence exists for an association of
O3 concentration with decreased sperm concentration and associations with reduced
birth weight and restricted fetal growth. Overall, the evidence is suggestive of a
causal relationship between long-term exposures to O3 and reproductive and
developmental effects.
These causal determinations are summarized in Table 2-1. along with the conclusions
from the previous NAAQS review. Special emphasis and additional details are
provided in Table 2-1 for respiratory health outcomes, for which there is the
strongest body of evidence.
Table 2-1 Summary of evidence from epidemiologic, controlled human
exposure, and animal toxicological studies on the health effects
associated with short- and long-term exposure to O3.
Health Outcome
Conclusions from 2006 O3 AQCD
Conclusions from this ISA
Short-Term Exposure to O3
Respiratory effects
The overall evidence supports a
causal relationship between acute
ambient O3 exposures and increased
respiratory morbidity outcomes.
Evidence integrated across controlled
human exposure, epidemiologic, and
toxicological studies and across the
spectrum of respiratory health endpoints
continues to demonstrate that there is a
causal relationship between short-term
O3 exposure and respiratory health
effects.
Lung function
Results from controlled human
exposure studies and animal
toxicological studies provide clear
evidence of causality for the
associations observed between acute
(< 24 h) O3 exposure and relatively
small, but statistically significant
declines in lung function observed in
numerous recent epidemiologic
studies. Declines in lung function are
particularly noted in children,
asthmatics, and adults who work or
exercise outdoors.
Recent controlled human exposure
studies demonstrate group mean
decreases in FEV-i in the range of 2 to
3% with 6.6 hour exposures to as low as
60 ppb O3. The collective body of
epidemiologic evidence demonstrates
associations between short-term ambient
O3 exposure and decrements in lung
function, particularly in children with
asthma, children, and adults who work or
exercise outdoors.
Airway
hyperresponsiveness
Evidence from human clinical and
animal toxicological studies clearly
indicate that acute exposure to O3
can induce airway hyperreactivity,
thus likely placing atopic asthmatics
at greater risk for more prolonged
bouts of breathing difficulties due to
airway constriction in response to
various airborne allergens or other
triggering stimuli.
A limited number of studies have
observed airway hyperresponsiveness in
rodents and guinea pigs after exposure to
less than 300 ppb O3. As previously
reported in the 2006 O3 AQCD, increased
airway responsiveness has been
demonstrated at 80 ppb in young, healthy
adults, and at 50 ppb in certain strains of
rats.
2-20
-------
Health Outcome
Conclusions from 2006 O3 AQCD
Conclusions from this ISA
Pulmonary
inflammation, injury
and oxidative stress
The extensive human clinical and
animal toxicological evidence,
together with the limited available
epidemiologic evidence, is clearly
indicative of a causal role for O3 in
inflammatory responses in the
airways.
Epidemiologic studies provided new
evidence for associations of ambient O^
with mediators of airway inflammation and
oxidative stress and indicate that higher
antioxidant levels may reduce pulmonary
inflammation associated with 63
exposure. Generally, these studies had
mean 8-h max Os concentrations less
than 73 ppb. Recent controlled human
exposure studies show O3-induced
inflammatory responses at 60 ppb, the
lowest concentration evaluated.
Respiratory
symptoms and
medication use
Young healthy adult subjects
exposed in clinical studies to Os
concentrations > 80 ppb for 6 to 8 h
during moderate exercise exhibit
symptoms of cough and pain on deep
inspiration. The epidemiologic
evidence shows significant
associations between acute exposure
to ambient O^ and increases in a
wide variety of respiratory symptoms
(e.g., cough, wheeze, production of
phlegm, and shortness of breath) and
medication use in asthmatic children.
The collective body of epidemiologic
evidence demonstrates positive
associations between short-term
exposure to ambient O3 and respiratory
symptoms (e.g., cough, wheeze, and
shortness of breath) in children with
asthma. Generally, these studies had
mean 8-h max O3 concentrations less
than 69 ppb.
Lung host defenses
Toxicological studies provided
extensive evidence that acute O3
exposures as low as 80 to 500 ppb
can cause increases in susceptibility
to infectious diseases due to
modulation of lung host defenses.
A single controlled human exposure
study found decrements in the ability
of alveolar macrophages to
phagocytize microorganisms upon
exposure to 80 to 100 ppb O3.
Recent controlled human exposure
studies demonstrate the increased
expression of cell surface markers and
alterations in sputum leukocyte markers
related to innate adaptive immunity with
short-term O3 exposures of 80-400 ppb.
Recent studies demonstrating altered
immune responses and natural killer cell
function build on prior evidence that O3
can affect multiple aspects of innate and
acquired immunity with short-term O3
exposures as low as 80 ppb.
Allergic and asthma
related responses
Previous toxicological evidence
indicated that O3 exposure skews
immune responses toward an allergic
phenotype, and enhances the
development and severity of asthma-
related responses such as AHR.
Recent controlled human exposure
studies demonstrate enhanced allergic
cytokine production in atopic individuals
and asthmatics, increased IgE receptors
in atopic asthmatics, and enhanced
markers of innate immunity and antigen
presentation in health subjects or atopic
asthmatics with short-term exposure to
80-400 ppb O3, all of which may enhance
allergy and/or asthma. Further evidence
for Os-induced allergic skewing is
provided by a few recent studies in
rodents using exposure concentrations as
low as 200 ppb.
2-21
-------
Health Outcome
Conclusions from 2006 O3 AQCD
Conclusions from this ISA
Respiratory
Hospital admissions,
ED visits, and
physician visits
Aggregate population time-series
studies observed that ambient O3
concentrations are positively and
robustly associated with respiratory-
related hospitalizations and asthma
ED visits during the warm season.
Consistent, positive associations of
ambient O3 with respiratory hospital
admissions and ED visits in the U.S.,
Europe, and Canada with supporting
evidence from single city studies.
Generally, these studies had mean
8-h max O3 concentrations less than
60 ppb.
Respiratory Mortality
Aggregate population time-series
studies specifically examining
mortality from respiratory causes
were limited in number and showed
inconsistent associations between
acute exposure to ambient O3
exposure and respiratory mortality.
Recent multicity time-series studies and a
multicontinent study consistently
demonstrated associations between
ambient O3 and respiratory-related
mortality visits across the U.S., Europe,
and Canada with supporting evidence
from single city studies. Generally, these
studies had mean 8-h max O3
concentrations less than 63 ppb.
Cardiovascular
effects
The limited evidence is highly
suggestive that O3 directly and/or
indirectly contributes to
cardiovascular-related morbidity, but
much remains to be done to more
fully substantiate the association.
The overall body of evidence across
disciplines indicates that there is likely to
be a causal relationship for short-term
exposures to O3 and cardiovascular
effects.
Central nervous
system effects
Toxicological studies report that
acute exposures to O3 are
associated with alterations in
neurotransmitters, motor activity,
short- and long-term memory, sleep
patterns, and histological signs of
neurodegeneration.
Together the evidence from studies of
short-term exposure to O3 is suggestive
of a causal relationship between Os
exposure and CNS effects.
Total Mortality
The evidence is highly suggestive
that O3 directly or indirectly
contributes to non-accidental and
cardiopulmonary-related mortality.
Taken together, the body of evidence
indicates that there is likely to be a
causal relationship between short-term
exposures to Os and total mortality.
Long-term Exposure to Os
Respiratory effects
The current evidence is suggestive
but inconclusive for respiratory health
effects from long-term O3 exposure.
Recent epidemiologic evidence,
combined with toxicological studies in
rodents and non-human primates,
provides biologically plausible evidence
that there is likely to be a causal
relationship between long-term
exposure to O3 and respiratory health
effects.
New onset asthma
No studies examining this outcome
were evaluated in the 2006 O3
AQCD.
Evidence that different genetic variants
(HMOX, GST, ARC), in combination with
O3 exposure, are related to new onset
asthma. These associations were
observed when subjects living in areas
where the mean annual 8-h max O3
concentration was 55.2 ppb, compared to
those who lived where it was 38.4 ppb.
2-22
-------
Health Outcome
Conclusions from 2006 O3 AQCD
Conclusions from this ISA
Asthma hospital
admissions
No studies examining this outcome
were evaluated in the 2006 Os
AQCD.
Chronic O3 exposure was related to first
childhood asthma hospital admissions in
a positive concentration-response
relationship. Generally, these studies had
mean annual 8-h max O3 concentrations
less than 41 ppb.
Pulmonary structure
and function
Epidemiologic studies observed that
reduced lung function growth in
children was associated with
seasonal exposure to Os', however,
cohort studies of annual or multiyear
Os exposure observed little clear
evidence for impacts of longer-term,
relatively low-level Os exposure on
lung function development in
children. Animal toxicological studies
reported chronic Os-induced
structural alterations, some of which
were irreversible, in several regions
of the respiratory tract including the
centriacinar region. Morphologic
evidence from studies using
exposure regimens that mimic
seasonal exposure patterns report
increased lung injury compared to
conventional chronic stable
exposures.
Evidence for pulmonary function effects is
inconclusive, with some new
epidemiologic studies observing positive
associations (mean annual 8-h max Os
concentrations less than 65 ppb).
Information from toxicological studies
indicates that long-term exposure during
development among infant monkeys
(500 ppb) and adult rodents (>120 ppb)
can result in irreversible morphological
changes in the lung, which in turn can
influence pulmonary function.
Pulmonary
inflammation, injury
and oxidative stress
Extensive human clinical and animal
toxicological evidence, together with
limited epidemiologic evidence
available, suggests a causal role for
Os in inflammatory responses in the
airways.
Several epidemiologic studies (mean
8-h max 63 concentrations less than
69 ppb) and toxicology studies (as low as
500 ppb) add to observations of
Os-induced inflammation and injury.
Lung host defenses
Toxicological studies provided
evidence that chronic Os exposure
as low as 100 ppb can cause
increases in susceptibility to
infectious diseases due to modulation
of lung host defenses, but do not
cause greater effects on infectivity
than short exposures.
Consistent with decrements in host
defenses observed in rodents exposed to
100 ppb O3, recent evidence
demonstrates a decreased ability to
respond to pathogenic signals in infant
monkeys exposed to 500 ppb Os.
Allergic responses
Limited epidemiologic evidence
supported an association between
ambient O3 and allergic symptoms.
Little if any information was available
from toxicological studies.
Evidence relates positive outcomes of
allergic response and Os exposure but
with variable strength for the effect
estimates; exposure to Os may increase
total IgE in adult asthmatics. Allergic
indicators in monkeys were increased by
exposure to Os concentrations of
500 ppb.
Respiratory mortality
Studies of cardio-pulmonary mortality
were insufficient to suggest a causal
relationship between chronic O3
exposure and increased risk for
mortality in humans.
A single study demonstrated that
exposure to 63 (long-term mean Os less
than 104 ppb) elevated the risk of death
from respiratory causes and this effect
was robust to the inclusion of
2-23
-------
Health Outcome
Conclusions from 2006 O3 AQCD
Conclusions from this ISA
Cardiovascular
Effects
No studies examining this outcome
were evaluated in the 2006 Os
AQCD.
The overall body of evidence across
disciplines is suggestive of a causal
relationship for long-term exposures to
Os and cardiovascular effects.
Reproductive and
developmental
effects
Limited evidence for a relationship
between air pollution and birth-
related health outcomes, including
mortality, premature births, low birth
weights, and birth defects, with little
evidence being found for Os effects.
Overall, the evidence is suggestive of a
causal relationship between long-term
exposures to Os and reproductive and
developmental effects.
Central nervous
system effects
Toxicological studies reported that
acute exposures to O3 are
associated with alterations in
neurotransmitters, motor activity,
short and long term memory, sleep
patterns, and histological signs of
neurodegeneration. Evidence
regarding chronic exposure and
neurobehavioral effects was not
available.
Together the evidence from studies of
long-term exposure to O3 is suggestive
of a causal relationship between Os
exposure and CNS effects.
Cancer
Little evidence for a relationship
between chronic O3 exposure and
increased risk of lung cancer.
Overall, the evidence is inadequate to
determine if a causal relationship
exists between ambient Os exposures
and cancer.
Total Mortality
There is little evidence to suggest a
causal relationship between chronic
Os exposure and increased risk for
mortality in humans.
Collectively, the evidence is suggestive
of a causal relationship between long-
term Os exposures and total mortality.
2.5.3 Integrated Synthesis of Evidence for Health Effects
This section integrates the evidence for respiratory and cardiovascular effects
(including mortality) across scientific disciplines and both short- and long-term
exposure periods. Here, the complete body of evidence from both previous and the
current NAAQS reviews is synthesized for the broad range of respiratory and
cardiovascular effects associated with exposure to O3.
2.5.3.1 Respiratory Effects
Building on evidence evaluated in previous O3 AQCDs, recent evidence confirms
and extends that O3 is associated with a broad range of respiratory effects, including
altered development of the respiratory tract. Recent animal toxicological studies of
long-term exposure to O3 occurring throughout various lifestages in monkeys,
beginning with prenatal and early life exposures, provide novel evidence for effects
on the development of the respiratory system, including ultrastructural changes in
bronchiole development, increased offspring airway hyper-reactivity (Section 7.4.8),
2-24
-------
as well as effects on the developing immune system. The strongest evidence for
O3-induced effects on the developing lung comes from a series of experiments using
infant rhesus monkeys episodically exposed to 500 ppb O3 for approximately
5 months, starting at one month of age. Functional changes in the conducting airways
of infant rhesus monkeys exposed to either O3 alone or O3 + antigen were
accompanied by a number of cellular and morphological changes. In addition to
these functional and cellular changes, substantial structural changes in the respiratory
tract were observed. Importantly, the O3-induced structural pathway changes
persisted after recovery in filtered air for six months after cessation of the O3
exposures. Exposure to O3 has also been associated with similar types of alterations
in pulmonary structure, including airways remodeling and pulmonary injury and
increased permeability, in all adult laboratory animal species studied, from rats to
monkeys (U.S. EPA. 1996a).
In addition to effects on the development and structure of the respiratory tract, there
is extensive evidence for the effects of short-term exposure to O3 on pulmonary
inflammation and oxidative stress. Previous evidence from controlled human
exposure studies indicated that O3 causes an inflammatory response in the lungs
(U.S. EPA, 1996a). This inflammatory response to O3 was detected after a single 1-h
exposure with exercise to O3 concentrations of 300 ppb; the increased levels of some
inflammatory cells and mediators persisted for at least 18 hours. Toxicological
studies provided additional evidence for increases in permeability and inflammation
in rabbits at levels as low as 100 ppb O3. Evidence summarized in the 2006 O3
AQCD demonstrated that inflammatory responses were observed subsequent to
6.6 hours O3 exposure to the lowest tested level of 80 ppb in healthy human adults,
while toxicological studies provided extensive evidence that short-term (1-3 hours)
O3 exposure in the range of 100-500 ppb could cause lung inflammatory responses.
The limited epidemiologic evidence reviewed in the 2006 O3 AQCD demonstrated
an association between short-term increases in ambient O3 concentration and airways
inflammation in children (1-h max O3 of approximately 100 ppb). Recent studies in
animals and in vitro models described inflammatory and injury responses mediated
by Toll-like receptors (e.g., TLR4, TLR2), receptors for TNF or IL-1, multiple
signaling pathways (e.g., p38, JNK, NFicB, MAPK/AP-1), and oxidative stress
(Section 6.2.3.3). Recent epidemiologic studies provide additional supporting
evidence by demonstrating associations of ambient O3 with mediators of airways
inflammation and oxidative stress.
The normal inflammatory response in lung tissue is part of host defense that aids in
removing microorganisms or particles that have reached the distal airways and
alveolar surface. The 1996 O3 AQCD concluded that short-term exposure to elevated
concentrations of O3 resulted in alterations in these host defense mechanisms in the
respiratory system. Specifically, toxicological studies of short-term exposures as low
as 100 ppb O3 for 2 hours were shown to decrease the ability of alveolar
macrophages to ingest particles, and short-term exposures as low as 80 ppb for
3 hours prevented mice from resisting infection with streptococcal bacteria and
resulted in infection-related mortality. Similarly, alveolar macrophages removed
from the lungs of human subjects after 6.6 hours of exposure to 80 and 100 ppb O3
2-25
-------
had decreased ability to ingest microorganisms, indicating some impairment of host
defense capability. These altered host defense mechanisms can lead to increased risk
of respiratory infections, which can often predispose individuals to developing
asthma when occurring in early life. Despite the strong toxicological evidence, in the
limited body of epidemiologic evidence, ambient O3 concentrations have not been
consistently associated with hospital admissions or ED visits for respiratory
infection, pneumonia, or influenza (Section 6.2.7.2 and Section 6.2.7.3).
The most commonly observed and strongest evidence for respiratory effects
associated with short-term exposure to O3 is transient decrements in pulmonary
function. Controlled human exposure studies reviewed in previous assessments
demonstrated O3-induced decrements in pulmonary function, characterized by
alterations in lung volumes and flow and airway resistance and responsiveness for
multihour exposures (up to 8 hours) to O3 concentrations as low as 80 ppb (U.S.
EPA, 1996a). A series of mobile laboratory studies of lung function and respiratory
symptoms reported pulmonary function decrements at mean ambient O3
concentrations of 140 ppb in exercising healthy adolescents and increased respiratory
symptoms and pulmonary function decrements at 150 ppb in heavily exercising
athletes and at 170 ppb in lightly exercising healthy and asthmatic subjects.
Epidemiologic and animal toxicological evidence is coherent with the results of the
controlled human exposure studies, both indicating decrements in lung function upon
O3 exposure. A combined statistical analysis of epidemiologic studies in children at
summer camp with particularly strong exposure assessment demonstrated decrements
in FEVi of 0.50 mL/ppb with an increase in previous hour O3 concentration. For
preadolescent children exposed to 120 ppb ambient O3, this estimated volume
decrease corresponded to an average decrement of 2.4-3.0% in FEVi. Key studies of
lung function (FEVi) measured before and after well-defined outdoor exercise events
in adults yielded concentration-response slopes of 0.40 and 1.35 mL/ppb ambient O3
after exposure for up to 1 hour. Animal toxicological studies reported similar
respiratory effects in rats at exposures as low as 200 ppb O3 for 3 hours. The 2006
O3 AQCD characterized the controlled human exposure and animal toxicological
studies as providing clear evidence of causality for the associations observed
between short-term (< 24 hours) increases in O3 concentration and relatively small,
but statistically significant declines in lung function observed in numerous recent
epidemiologic studies. In epidemiologic studies, declines in lung function were
particularly noted in children with and without asthma, and adults who work or
exercise outdoors.
Recent controlled human exposure studies examined lower concentration O3
exposures (40-80 ppb) and demonstrated thatFEVj, respiratory symptoms, and
inflammatory responses were affected by O3 exposures of 6.6 hours as low as 60 to
70 ppb (Section 6.2.1.1 and Section 6.2.3.1). These studies demonstrated average
O3-induced decreases in FEVi in the range of 2.8 to 3.6% with O3 exposures to
60 ppb for 6.6 hours. Further, in the controlled human exposure studies evaluating
effects of 60 ppb O3, on average, 10% of the exposed individuals experienced >10%
FEVi decrements following 6.6 hours of exposure. Considerable intersubject
variability has also been reported in studies at higher exposure concentrations
2-26
-------
(> 70 ppb) with some subjects experiencing considerably greater decrements than
average. Recent epidemiologic studies provide greater insight into individual- and
population-level factors that can increase for the risk of O3-associated respiratory
morbidity. In addition to lung function decrements consistently reported in healthy
children at summer camp, O3-associated decreases in lung function were consistently
observed in epidemiologic studies that included potentially at-risk populations
(e.g., individuals with asthma with concurrent respiratory infection, older adults with
AHR or elevated body mass index, or groups with diminished antioxidant capacity).
Exposure to O3 can also result in respiratory symptoms (e.g., coughing, wheezing,
shortness of breath). The 1996 O3 AQCD identified an association between
respiratory symptoms and increasing ambient O3, particularly among children with
asthma. In the 2006 O3 AQCD, symptoms of cough and pain on deep inspiration
were well documented in young healthy adult subjects after exposure of > 80 ppb O3
for 6-8 hours during moderate exercise. Limited data suggested an increase in
respiratory symptoms down to 60 ppb. More recently, these effects have been
observed at 70 ppb in healthy adults. Controlled human exposure studies of healthy
adults, have also reported an increased incidence of cough with O3 exposures as low
as 120 ppb and 1-3 hours in duration with very heavy exercise. The controlled human
exposure studies also demonstrated lesser respiratory symptom responses in children
and older adults relative to young healthy adults. Cumulative epidemiologic evidence
adds to the findings from controlled human exposure studies for healthy adults by
demonstrating the effects of ambient O3 exposure on respiratory symptoms in
children with asthma. Increases in ambient O3 concentration were associated with a
wide variety of respiratory symptoms (e.g., cough, wheeze, and shortness of breath)
in children with asthma. Epidemiologic studies also indicated that short-term
increases in O3 concentration are likely associated with increased asthma medication
use in children with asthma. Additionally, epidemiologic studies provide evidence
for an association between long-term exposure to O3 and respiratory symptoms
(Section 7.2).
Ozone exposure has been shown to result in both specific and non-specific airway
hyperresponsiveness (AHR). Increased AHR is an important consequence of
exposure to O3 because its presence represents a change in airway smooth muscle
reactivity and implies that the airways are predisposed to narrowing on inhalation of
a variety of stimuli (e.g., specific allergens, SO2, cold air). Specifically, short-term (2
or 3 hours) exposure to 250 or 400 ppb O3 was found to cause increases in AHR in
response to allergen challenges among allergic asthmatic subjects who
characteristically already had somewhat increased AHR at baseline. Increased non-
specific AHR has been demonstrated in healthy young adults down to 80 ppb O3
following 6.6 hours of exposure during moderate exercise. While AHR has not been
widely examined in epidemiologic studies, findings for O3-induced increases in AHR
in controlled human exposure (Section 6.2.2.1) and toxicological (Section 6.2.2.2)
studies provide biological plausibility for associations observed between increases in
ambient O3 concentration and increases in respiratory symptoms in subjects with
asthma.
2-27
-------
In addition to asthma exacerbations, recent epidemiologic evidence has indicated that
long-term ambient O3 concentrations can contribute to new onset asthma (Section
7.2.1, Table 7-2). The new epidemiologic evidence base consists of studies using a
variety of designs and analysis methods evaluating the relationship between long-
term annual measures of exposure to ambient O3 and measures of respiratory
morbidity. Studies from two California cohorts have provided evidence for different
variants in genes related to oxidative or nitrosative stress (e.g., HMOX, GSTs, ARG)
that, depending on community long-term O3 concentrations, are related to new onset
asthma. These cohorts provide evidence that extends beyond the association of short-
term exposure to O3 and asthma exacerbations to suggest that long-term exposure to
O3 may play a role in the development of the disease and contribute to incident cases
of asthma.
The frequency of ED visits and hospital admissions due to respiratory symptoms,
asthma exacerbations and other respiratory diseases is associated with short- and
long-term exposure to ambient O3 concentrations. Summertime daily hospital
admissions for respiratory causes in various locations of eastern North America were
consistently associated with ambient concentrations of O3 in studies reviewed in the
1996 O3 AQCD. This association remained even with examination of only
concentrations below 120 ppb O3. The 2006 O3 AQCD concluded that aggregate
population time-series studies demonstrate a positive and robust association between
ambient O3 concentrations and respiratory-related hospitalizations and asthma ED
visits during the warm season. Recent epidemiologic time-series studies that include
additional multicity studies and a multicontinent study further demonstrate that short-
term exposures to ambient O3 concentrations are consistently associated with
increases in respiratory hospital admissions and ED visits specifically during the
warm/summer months across a range of O3 concentrations (Section 6.2.7). There is
also recent evidence for an association between respiratory hospital admissions and
long-term exposure to O3 (Section 7.2.2).
Finally, O3 exposure can contribute to death from respiratory causes. Recent
evidence from several multicity studies and a multicontinent study demonstrate
consistent positive associations between short-term exposure to ambient O3
concentrations and increases in respiratory mortality (Section 6.2.8). Similarly, a
study of long-term exposure to ambient O3 concentrations also demonstrated an
association between O3 and increases in respiratory mortality (Section 7.7.1).
Evidence from these recent mortality studies is consistent and coherent with the
evidence from epidemiologic, controlled human exposure, and animal toxicological
studies for the effects of short- and long-term exposure to O3 on respiratory effects.
Additionally, the evidence for respiratory morbidity after short- and long-term
exposure provides biological plausibility for mortality due to respiratory disease.
2-28
-------
2.5.3.2 Cardiovascular Effects
There is an emerging body of animal toxicological evidence demonstrating that
short-term exposure to O3 can lead to autonomic nervous system alterations (in heart
rate and/or heart rate variability) and suggesting that proinflammatory signals may
mediate cardiovascular effects. Interactions of O3 with respiratory tract components
result in secondary oxidation products and inflammatory mediators that have the
potential to penetrate the epithelial barrier and to initiate toxic effects systemically.
In addition, animal toxicological studies of long-term exposure to O3 provide
evidence of enhanced atherosclerosis and I/R injury, corresponding with
development of a systemic oxidative, proinflammatory environment.
Evidence from controlled human exposure studies also demonstrates cardiovascular
effects in response to short-term O3 exposure and provides some coherence with
evidence from animal toxicology studies. Controlled human exposure studies support
the animal toxicological studies by demonstrating O^-induced effects on blood
biomarkers of systemic inflammation and oxidative stress, preliminary evidence for
O 3-induced modulation of the autonomic nervous system, as well as changes in
biomarkers that can indicate aprothrombogenic response to O3. However,
epidemiologic studies evaluating cardiovascular morbidity and short- and long-term
exposure to O3 do not provide consistent evidence for an association. This is evident
by the multiple studies that examined the association between (1) short- and long-
term O3 concentrations and cardiovascular-related hospital admissions and ED visits,
and (2) cardiovascular disease-related biomarkers, and reported inconsistent results.
When examining mortality due to cardiovascular disease, epidemiologic studies
consistently observe positive associations with short-term exposure to O3.
Additionally, there is some evidence for an association between long-term exposure
to O3 and mortality. However, the association between long-term ambient O3
concentrations and cardiovascular mortality can be confounded by other pollutants as
evident by a study of cardiovascular mortality that reported no association after
adjustment for PM2.s concentrations.
Overall, animal toxicological studies demonstrate O3-induced cardiovascular effects,
and support the strong body of evidence indicating O3-induced cardiovascular
mortality. Animal toxicological and controlled human exposure studies provide
evidence for biologically plausible mechanisms underlying these O3-induced
cardiovascular effects. However, a lack of coherence with epidemiologic studies of
cardiovascular morbidity remains an important uncertainty.
2.5.4 Policy Relevant Considerations
This ISA summarizes and integrates the available scientific evidence that best
informs consideration of the policy-relevant questions that frame this assessment,
presented in the Integrated Review Plan (U.S. EPA. 20lid). This includes
considering whether the available body of scientific evidence supports or calls into
2-29
-------
question the scientific conclusions reached in the last review regarding health effects
related to exposure to O3, with particular emphasis on exposures and health risks
among populations potentially at increased risk. Additional policy relevant
considerations include how the scientific information, when available, informs
decisions regarding the basic elements of the NAAQS: indicator, averaging time,
level, and form.
2.5.4.1 Populations Potentially at Increased Risk
Studies were reviewed to identify populations that are at increased risk for O3-related
health effects. These studies have investigated factors that can cause populations to
be at increased risk for O3-related health effects by conducting stratified
epidemiologic analyses; by examining individuals with an underlying health
condition, genetic polymorphism, or categorized by age, race, or sex in controlled
human exposure studies; or by developing animal models that mimic the
pathophysiological conditions associated with a health effect. These studies have
identified a multitude of factors that could potentially contribute to whether a
population is at increased risk for O3-related health effects.
The populations identified in Chapter 8 that were examined for their potential for
increased risk of O3-related health effects are listed in Table 8-6 and are classified as
providing adequate, suggestive, inadequate, or no evidence of being an at-risk factor.
The factors that have adequate evidence to be classified as an at-risk factor for
O3-related health effects are individuals with asthma, younger and older age groups,
individuals with reduced intake of certain nutrients (i.e., vitamins C and E), and
outdoor workers, based on consistency in findings across studies and evidence of
coherence in results from different scientific disciplines. Asthma as a factor affecting
risk was supported by controlled human exposure and toxicological studies, as well
as some evidence from epidemiologic studies. Generally, studies comparing age
groups also reported greater associations for respiratory hospital admissions and ED
visits among children than for adults. Biological plausibility for this increased risk is
supported by toxicological and controlled human exposure studies. Also, children
have higher exposure and dose due to increased time spent outdoors and ventilation
rate, and childrens' respiratory systems are also still undergoing lung growth. Most
studies comparing age groups reported greater effects of short-term O3 exposure on
mortality among older adults, although studies of other health outcomes had
inconsistent findings regarding whether older adults were at increased risk. Multiple
epidemiologic, controlled human exposure, and toxicological studies reported that
diets lower in vitamins E and C are associated with increased risk of O3-related
health effects. Previous studies have shown that increased exposure to O3 due to
outdoor work leads to increased risk of O3-related health effects and it is clear that
outdoor workers have higher exposures, and greater internal doses, of O3, which may
lead to increased risk of O3-related health effects.
2-30
-------
Other potential factors [genetic variants (such as those in GSTM1, HMOX-1, NQO1,
and TNF-a), obesity, sex, and SES] provided some suggestive evidence of increased
risk, but further investigation is needed. Similarly, many factors had inadequate
evidence to determine if they increased the risk of O3-related health effects,
including influenza/infection, COPD, CVD, diabetes, hyperthyroidism, smoking,
race/ethnicity, and air conditioning use.
2.5.4.2 Exposure Metrics in Epidemiologic Studies
Some epidemiologic studies have conducted analyses between O3 concentration and
health effects (i.e., mortality, respiratory or cardiovascular) using various exposure
metrics (i.e., 1-h max, 8-h max, and 24-h avg). No studies of long-term exposure
(i.e., months to years) to O3 have compared the use of different exposure metrics on
risk estimation.
Among time-series studies, the limited evidence suggests comparable risk estimates
across exposure metrics with some evidence for smaller O3 risk estimates when
using a 24-hour average exposure metric. Several panel studies examined whether
associations of lung function and respiratory symptoms varied depending on the O3
exposure metric used. Although differences in effect estimates across exposure
metrics were found within some studies, collectively, there was no indication that the
consistency or magnitude of the observed association was stronger for a particular O3
exposure metric. Several studies examining lung function demonstrated that this was
true among populations with and without increased outdoor exposures. It is important
to note in these studies, the degree of exposure measurement error associated with
use of central site ambient O3 concentrations may vary among O3 averaging times,
depending on time spent outdoors. Among studies that examined associations of
multiple respiratory symptoms in children with multiple O3 exposure metrics, most
did not find higher odds ratios for any particular exposure metric. Overall, the
evidence from time-series and panel epidemiologic studies does not indicate that one
exposure metric is more consistently or strongly associated with mortality or
respiratory-related health effects.
2.5.4.3 Lag Structure in Epidemiologic Studies
Epidemiologic studies have attempted to identify the time-frame in which exposure
to O3 can impart a health effect. The time period between O3 exposure and health
effects can potentially be influenced by a multitude of factors, such as age or
existence of pre-existing diseases. Different lag times have been evaluated for
specific health outcomes.
The epidemiologic evidence evaluated in the 2006 O3 AQCD indicated that one of
the remaining uncertainties in characterizing the O3-mortality relationship was
identifying the appropriate lag structure (e.g., single-day lags versus distributed lag
2-31
-------
model). An examination of lag times used in the epidemiologic studies evaluated in
this assessment can provide further insight on the characterization of the relationship
between O3 exposure and morbidity and mortality outcomes from epidemiologic
studies.
The majority of epidemiologic studies that focused on the association between short-
term O3 exposure and mortality (i.e., all-cause, respiratory and cardiovascular)
examined the average of multiday lags with some studies examining single-day lags.
Across a range of multiday lags (i.e., average of 0-1 to 0-6 days), the studies
evaluated consistently demonstrate that the O3 effects on mortality occur within a
few days of exposure (Figure 6-27).
Epidemiologic studies of lung function, respiratory symptoms, and biological
markers of airway inflammation and oxidative stress examined associations with
single-day ambient O3 concentrations (using various averaging times) lagged from 0
to 7 days as well as concentrations averaged over 2 to 19 days. Lags of 0 and 1 day
ambient O3 concentrations were associated with decreases in lung function and
increases in respiratory symptoms, airway inflammation, and oxidative stress.
Additionally, several studies found that multiday averages of O3 concentration were
associated with these endpoints, indicating that not only single day, but exposures
accumulated over several days led to a respiratory health effect. In studies of
respiratory hospital admissions and ED visits, investigators either examined the lag
structure of associations by including both single-day and the average of multiday
lags, or selecting lags a priori. The collective evidence indicates a rather immediate
response within the first few days of O3 exposure (i.e., for lags days averaged at 0-1,
0-2, and 0-3 days) for hospital admissions and ED visits for all respiratory outcomes,
asthma, and chronic obstructive pulmonary disease in all-year and seasonal analyses.
2.5.4.4 Ozone Concentration-Response Relationship
An important consideration in characterizing the O3-morbidity and mortality
association is whether the concentration-response (C-R) relationship is linear across
the full concentration range that is encountered or if there are concentration ranges
where there are departures from linearity (i.e., nonlinearity). In this ISA studies have
been identified that attempt to characterize the shape of the O3 C-R curve along with
possible O3 "thresholds" (i.e., O3 concentrations which must be exceeded in order to
elicit an observable health response). The epidemiologic studies that examined the
shape of the C-R curve and the potential presence of a threshold have indicated a
generally linear C-R function with no indication of a threshold in analyses that have
examined 8-h max and 24-h avg O3 concentrations. However, there is less certainty
in the shape of the C-R curve at the lower end of the distribution of O3
concentrations (below NA background concentrations, 29-40 ppb) due to the low
density of data in this range.
Controlled human exposure studies have provided strong and quantifiable C-R data
on the human health effects of O3. The magnitude of respiratory effects in these
2-32
-------
studies is generally a function of O3 exposure, i.e., the product of concentration (C),
minute ventilation (VE), and exposure duration. Several studies provide evidence for
a smooth C-R curve in young healthy adults exposed during moderate exercise for
6.6 hours to O3 concentrations between 40 and 120 ppb (Figure 6-1). It is difficult to
characterize the C-R relationship at and below 40 ppb due to uncertainty associated
with the sparse data at these lower concentrations.
Although relatively few epidemiologic studies have examined the O3-health effects
C-R relationship, the C-R relationship has been examined across multiple health
endpoints and exposure durations. Some studies of populations engaged in outdoor
activity found that associations between O3 and lung function decrements persisted at
lower O3 concentrations with some studies showing larger negative associations in
analyses limited to lower O3 concentrations (e.g., 60-80 ppb; Table 6-6) and shorter
exposure durations (i.e., in the range of 30 minutes to less than 8 hours; Table 6-6).
A study examining the C-R relationship between short-term O3 exposure and
pediatric asthma ED visits found no evidence of a threshold with a linear relationship
evident down to 8-h max O3 concentrations as low as 30 ppb (Figure 6-17). In an
additional study, authors used a smooth function while also accounting for the
potential confounding effects of PM2.s, to examine whether the shape of the C-R
curve for short-term exposure to O3 and asthma hospital admissions is linear. When
comparing the curve to a linear fit, the authors found that the linear fit is a reasonable
approximation of the C-R relationship between O3 and asthma hospital admissions in
the mid-range of the data though it can be seen that there is greater uncertainty at the
lower end of the distribution of ambient O3 concentrations, generally below 20 ppb
(Figure 6-16) due to sparse data at these lower concentrations.
Several recent studies applied a variety of statistical approaches to examine the shape
of the O3-mortality C-R relationship and existence of a threshold (Section 6.6.2.4).
These studies suggest that the shape of the O3-mortality C-R curve is linear across
the range of O3 concentrations though uncertainty in the relationship increases at the
lower end of the distribution (Figure 6-36). Generally, the epidemiologic studies that
examined the O3-mortality C-R relationship do not provide evidence for the
existence of a threshold within the range of 24-h average (24-h avg) O3
concentrations most commonly observed in the U.S. during the O3 season (i.e., above
20 ppb). However, the evaluation of the C-R relationship for short-term exposure to
O3 and mortality is difficult due to the evidence from multicity studies indicating
heterogeneity in O3-mortality associations across regions of the United States.
In addition, there are numerous issues that can influence the shape of the
O3-mortality C-R relationship that need to be taken into consideration including:
multiday effects (distributed lags), and potential adaptation and mortality
displacement (i.e., hastening of death by a short period). Additionally, given the
effect modifiers identified in mortality analyses that are also expected to vary
regionally (e.g., temperature, air conditioning prevalence), a national or combined
analysis may not be appropriate to identify whether a threshold exists in the
O3-mortality C-R relationship.
2-33
-------
In addition, the C-R relationship of long-term exposure to O3 and birth outcomes has
been evaluated. Evidence from the southern California Children's Health Study
identified a C-R relationship of birth weight with 24-h avg O3 concentrations
averaged over the entire pregnancy that was clearest above the 30 ppb level
(Figure 7-4).
Generally, both short- and long-term exposure studies indicate a linear, no threshold
C-R relationship when examining the association between O3 exposure and multiple
health effects across the range of 8-h max and 24-h avg O3 concentrations most
commonly observed in the U.S. during the O3 season (i.e., greater than 20 ppb).
However, evidence from studies of respiratory health effects and mortality indicates
less certainty in the shape of the C-R curve at the lower end of the distribution of O3
data, which corresponds to 8-h max and 24-h avg O3 concentrations generally below
20 ppb.
2.5.4.5 Regional Heterogeneity in Risk Estimates
Multicity epidemiologic studies that have examined the relationship between short-
term O3 exposures and mortality have provided evidence of city-to-city and regional
heterogeneity in O3-mortality risk estimates. A possible explanation for this
heterogeneity may be differences in community characteristics (individual- or
community-level) across cities that could modify the O3 effect (e.g., activity patterns,
housing type and age distribution, prevalence and use of air conditioning). Another
possible explanation for the observed heterogeneity could be effect modification by
concentrations of other air pollutants or interactions with temperature or other
meteorological factors that vary regionally in the U.S.
An examination of community characteristics measured at the individual level that
may contribute to the observed heterogeneity in O3-mortality risk estimates indicates
increased risk in older adults (i.e., > 65 years of age), women, African American
individuals, individuals with pre-existing diseases/conditions (e.g., diabetes, atrial
fibrillation), and lower SES. Furthermore, studies have examined community
characteristics measured at the community level and found that higher O3-mortality
risk estimates were associated with higher: percent unemployment, fraction of the
population Black/African-American, percent of the population that take public
transportation to work; and with lower: temperatures and percent of households with
central air conditioning. There is also evidence of greater effects in cities with lower
mean O3 concentrations. Additionally, there is evidence of increased risk of
O3-related mortality as percentage unemployed increases and a reduction in
O3-related mortality as mean temperature increased (i.e., a surrogate for air
conditioning rate) in the United States. The lack of a consistent reduction in O3-risk
estimates in cities with a higher percentage of central air conditioning across health
outcomes complicates the interpretation of the potential modifying effects of air
conditioning use.
2-34
-------
Overall, the epidemiologic studies that have examined the city-to-city and regional
heterogeneity observed in multicity studies have identified a variety of factors that
may modify the O3-mortality or -respiratory hospital admission relationship. Some
studies have also examined the correlation with other air pollutants or the potential
interactive effects between O3 and temperature to explain city-to-city heterogeneity
in Os-mortality risk estimates. This includes evidence that O3-mortality risk
estimates in the U.S. varied by mean SO2 concentrations, the ratio between mean
NO2/PMio concentrations, and temperatures. However, studies have not consistently
identified specific community characteristics that explain the observed heterogeneity.
2.6 Integration of Effects on Vegetation and Ecosystems
Chapter 9 presents the most policy-relevant information related to this review of the
NAAQS for the welfare effects of O3 on vegetation and ecosystems. This section
integrates the key findings from the disciplines evaluated in this assessment of the O3
scientific literature, which includes plant physiology, whole plant biology,
ecosystems, and exposure-response.
Overall, exposure to O3 is causally related or likely to be causally related to effects
observed on vegetation and ecosystems. These effects are observed across the entire
continuum of biological organization; from the cellular and subcellular level to the
whole plant level, and up to ecosystem-level processes. Furthermore, there is
evidence that the effects observed across this continuum are related to one another;
effects of O3 at lower levels of organization, such as the leaf of an individual plant,
can result in effects at higher levels. Ozone enters leaves through stomata, and can
alter stomatal conductance and disrupt CO2 fixation (Section 9.3). These effects can
change rates of leaf gas exchange, growth and reproduction at the individual plant
level and result in changes in ecosystems, such as productivity, C storage, water
cycling, nutrient cycling, and community composition (Section 9.4). Figure 2-2 is a
simplified illustrative diagram of the major endpoints that O3 may affect in
vegetation and ecosystems.
The framework for causal determinations (see Preamble) has been applied to the
body of scientific evidence to examine effects attributed to O3 exposure (Table 2-2).
The summary below provides brief integrated summaries of the evidence that
supports the causal determinations. The detailed discussion of the underlying
evidence used to formulate each causal determination can be found in Chapter 9.
This summary ends with a short discussion of policy relevant considerations.
2.6.1 Visible Foliar Injury
Visible foliar injury resulting from exposure to O3 has been well characterized and
documented over several decades of research on many tree, shrub, herbaceous, and
crop species (U.S. EPA. 2006b. 1996b. 1984. 1978a) (Section 9.4.2). Ozone-induced
2-35
-------
visible foliar injury symptoms on certain bioindicator plant species are considered
diagnostic as they have been verified experimentally in exposure-response studies,
using exposure methodologies such as continuous stirred tank reactors (CSTRs),
open-top chambers (OTCs), and free-air fumigation. Experimental evidence has
clearly established a consistent association of visible injury with O3 exposure, with
greater exposure often resulting in greater and more prevalent injury. Since
publication of the 2006 O3 AQCD, the results of several multiple-year field surveys
of Os-induced visible foliar injury at National Wildlife Refuges in Maine, Michigan,
New Jersey, and South Carolina have been published. New sensitive species showing
visible foliar injury continue to be identified from field surveys and verified in
controlled exposure studies.
Effects of Ozone Exposure
Leaf metabolism & physiology
•Antioxidant metabolism up-regulated
•Decreased photosynthesis
•Decreased stomatal conductance
or sluggish stomatal response
Leaves & canopy
-Visible leaf injury
-Altered leaf senescence
•Altered leaf chemical composition
Plant growth (Fig 9-8)
•Decreased biomass accumulation
•Altered reproduction
•Altered carbon allocation
•Altered crop quality
Belowground processes (Fig 9-8)
•Altered litter production & decomposition
•Altered soil carbon & nutrient cycling
•Altered soil fauna & microbial communities
Ecosystem services
•Decreased productivity
•Decreased C sequestration
•Altered water cycling (Fig 9-7)
•Altered community composition
(i.e., plant, insect & microbe)
Figure 2-2 An illustrative diagram of the major endpoints that O3 may affect in
plants and ecosystems.
2-36
-------
Table 2-2 Summary of O3 causal determinations for vegetation and
ecosystem effects.
Vegetation and
Ecosystem Effects
Conclusions from 2006 O3 AQCD
Conclusions from this
ISA
Visible Foliar Injury
Effects on Vegetation
Data published since the 1996 O3 AQCD strengthen
previous conclusions that there is strong evidence that
current ambient O3 concentrations cause impaired aesthetic
quality of many native plants and trees by increasing foliar
injury.
Causal Relationship
Reduced Vegetation
Growth
Data published since the 1996 O3 AQCD strengthen
previous conclusions that there is strong evidence that
current ambient O3 concentrations cause decreased growth
and biomass accumulation in annual, perennial and woody
plants, including agronomic crops, annuals, shrubs,
grasses, and trees.
Causal Relationship
Reduced Productivity in
Terrestrial Ecosystems
There is evidence that O3 is an important stressor of
ecosystems and that the effects of O3 on individual plants
and processes are scaled up through the ecosystem,
affecting net primary productivity.
Causal Relationship
Reduced Carbon
(C) Sequestration in
Terrestrial Ecosystems
Limited studies from the 2006 O3 AQCD.
Likely to be a
Causal Relationship
Reduced Yield and
Quality of Agricultural
Crops
Data published since the 1996 O3 AQCD strengthen
previous conclusions that there is strong evidence that
current ambient O3 concentrations cause decreased yield
and/or nutritive quality in a large number of agronomic and
forage crops.
Causal Relationship
Alteration of Terrestrial
Ecosystem Water
Cycling
Ecosystem water quantity may be affected by O3 exposure
at the landscape level.
Likely to be a
Causal Relationship
Alteration of Below-
ground Biogeochemical
Cycles
Ozone-sensitive species have well known responses to O3
exposure, including altered C allocation to below-ground
tissues; and also altered rates of leaf and root production,
turnover, and decomposition. These shifts can affect overall
C loss and nitrogen (N) loss from the ecosystem in terms of
respired C, and leached aqueous dissolved organic and
inorganic C and N.
Causal Relationship
Alteration of Terrestrial
Community Composition
Ozone may be affecting above- and below -ground
community composition through impacts on both growth and
reproduction. Significant changes in plant community
composition resulting directly from O3 exposure have been
demonstrated.
Likely to be a
Causal Relationship
The use of biological indicators in field surveys to detect phytotoxic levels of O3 is a
longstanding and effective methodology. The USDA Forest Service through the
Forest Health Monitoring (FHM) Program (1990-2001) and currently the Forest
Inventory and Analysis (FIA) Program has been collecting data regarding the
incidence and severity of visible foliar injury on a variety of O3 sensitive plant
species throughout the United States. The network has provided evidence that O3
concentrations were high enough to induce visible symptoms on sensitive vegetation.
2-37
-------
From repeated observations and measurements made over a number of years, specific
geographical patterns of visible O3 injury symptoms can be identified. In addition, a
study assessed the risk of O3-induced visible foliar injury on bioindicator plants in
244 national parks in support of the National Park Service's Vital Signs Monitoring
Network. The results of the study demonstrated that the estimated risk of visible
foliar injury was high in 65 parks (27%), moderate in 46 parks (19%), and low in 131
parks (54%). Some of the well-known parks with a high risk of O3-induced visible
foliar injury include Gettysburg, Valley Forge, Delaware Water Gap, Cape Cod, Fire
Island, Antietam, Harpers Ferry, Manassas, Wolf Trap Farm Park, Mammoth Cave,
Shiloh, Sleeping Bear Dunes, Great Smoky Mountains, Joshua Tree, Sequoia and
Kings Canyon, and Yosemite. Overall, evidence is sufficient to conclude that there is
a causal relationship between ambient O3 exposure and the occurrence of
Os-induced visible foliar injury on sensitive vegetation across the U.S.
2.6.2 Growth, Productivity, Carbon Storage and Agriculture
Ambient O3 concentrations have long been known to cause decreases in
photosynthetic rates and plant growth. The O3-induced damages at the plant scale
may translate to damages at the stand, then ecosystem scales, and cause changes in
productivity and C storage. The effects of O3 exposure on photosynthesis, growth,
biomass allocation, ecosystem production, and ecosystem C sequestration were
reviewed for the natural ecosystems, and crop productivity and crop quality were
reviewed for the agricultural ecosystems.
2.6.2.1 Natural Ecosystems
The previous O3 AQCDs concluded that there is strong and consistent evidence that
ambient concentrations of O3 decrease plant photosynthesis and growth in numerous
plant species across the United States. Studies published since the last review
continue to support that conclusion (Section 9.4.3.1). Recent studies, based on the
Aspen free-air carbon-dioxide/ozone enrichment (FACE) experiment, found that O3
caused reductions in total biomass relative to the control in aspen, paper birch, and
sugar maple communities during the first seven years of stand development. Overall,
the studies at the Aspen FACE experiment were consistent with the open-top
chamber (OTC) studies that were the foundation of previous O3 NAAQS reviews.
These results strengthen the understanding of O3 effects on forests and demonstrate
the relevance of the knowledge gained from trees grown in OTC studies.
A set of meta-analyses assessed the effects of O3 on plant photosynthesis and growth
across different species and fumigation methods (such as OTC and FACE). Those
studies reported that current O3 concentrations in the northern hemisphere are
decreasing photosynthesis (~11%) across tree species, and the decreases in
photosynthesis are consistent with cumulative uptake of O3 into the leaf. The current
ambient O3 concentrations (~40 ppb averaged across all hours of exposure)
2-38
-------
decreased annual total biomass growth of forest species by an average of 7%, with
potentially greater decreases (11-17%) with elevated O3 exposures (Section 9.4.3.1).
The meta-analyses further confirmed that reduction of plant photosynthesis and
growth under O3 exposure are coherent across numerous species and various
experimental techniques.
Studies during recent decades have also demonstrated O3 alters biomass allocation
and plant reproduction (Section 9.4.3). Recent meta-analyses have generally
indicated that O3 reduced C allocated to roots. Several recent studies published since
the 2006 O3 AQCD further demonstrate that O3 altered reproductive processes, such
as timing of flowering, number of flowers, fruits and seeds, in herbaceous and woody
plant species. However, a knowledge gap still exists pertaining to the exact
mechanism of the responses of reproductive processes to O3 exposure
(Section 9.4.3.3).
Studies at the leaf and plant scales show that O3 decreased photosynthesis and plant
growth, providing coherence and biological plausibility for the reported decreases in
ecosystem productivity. During the previous NAAQS reviews, there were very few
studies that investigated the effect of O3 exposure on ecosystem productivity and
C sequestration. Recent studies from long-term FACE experiments and ecosystem
models provided evidence of the association of O3 exposure and reduced
productivity at the ecosystem scale. Elevated O3 reduced stand biomass at Aspen
FACE after 7 years of O3 exposure, and annual volume growth at the Kranzberg
Forest in Germany. Results across different ecosystem models were consistent with
the FACE experimental evidence, which showed that O3 reduced ecosystem
productivity (Section 9.4.3.4). In addition to primary productivity, other indicators
such as net ecosystem productivity (NEP), net ecosystem CO2 exchange (NEE) and
C sequestration were often assessed by model studies. Model simulations
consistently found that O3 exposure caused negative impacts on these indicators
(Section 9.4.3.4, Table 9-3), but the severity of these impacts was influenced by
multiple interactions of biological and environmental factors. The suppression of
ecosystem C sinks results in more CO2 accumulation in the atmosphere. A recent
study suggested that the indirect radiative forcing caused by O3 exposure through
lowering the ecosystem C sink could have an even greater impact on global warming
than the direct radiative forcing of O3.
Although O3 generally causes negative effects on ecosystem productivity, the
magnitude of the response varies among plant communities (Section 9.4.3.4). For
example, O3 had little impact on white fir, but greatly reduced growth of ponderosa
pine in southern California. Ozone decreased net primary production (NPP) of most
forest types in the Mid-Atlantic region, but had small impacts on spruce-fir forest.
Ozone could also affect regional C budgets through interacting with multiple factors,
such as N deposition, elevated CO2 and land use history. Model simulations
suggested that O3 partially offset the growth stimulation caused by elevated CO2 and
N deposition in both Northeast- and Mid-Atlantic-region forest ecosystems of the
U.S.
2-39
-------
Overall, evidence is sufficient to conclude that there is a causal relationship
between ambient O3 exposure and reduced native plant growth and
productivity, and a likely causal relationship between O3 exposure and reduced
carbon sequestration in terrestrial ecosystems.
2.6.2.2 Agricultural Crops
The detrimental effect of O3 on crop production has been recognized since the 1960s
and a large body of research has subsequently stemmed from those initial findings.
Previous O3 AQCDs have extensively reviewed this body of literature. Current O3
concentrations across the U.S. are high enough to cause yield loss for a variety of
agricultural crops including, but not limited to, soybean, wheat, potato, watermelon,
beans, turnip, onion, lettuce, and tomato (Section 9.4.4.1). Continued increases in O3
concentration may further decrease yield in these sensitive crops. Despite the well-
documented yield losses due to increasing O3 concentration, there is still a
knowledge gap pertaining to the exact mechanism of O3-induced yield loss. Research
has linked increasing O3 concentration to decreased photosynthetic rates and
accelerated senescence, which are related to yield.
In addition, recent research has highlighted the effects of O3 on crop quality.
Increasing O3 concentration decreases nutritive quality of grasses, decreases macro-
and micro-nutrient concentrations in fruits and vegetable crops, and decreases cotton
fiber quality. These areas of research require further investigation to determine the
mechanism and dose-responses (Section 9.4.4.2).
During the previous NAAQS reviews, there were very few studies that estimated O3
impacts on crop yields at large geographical scales (i.e., regional, national or global).
Recent modeling studies found that O3 generally reduced crop yield, but the impacts
varied across regions and crop species (Section 9.4.4.1). For example, the largest
O3-induced crop yield losses occurred in high-production areas exposed to high O3
concentrations, such as the Midwest and the Mississippi Valley regions of the United
States. Among crop species, the estimated yield loss for wheat and soybean were
higher than rice and maize. Satellite and ground-based O3 measurements have been
used to assess yield loss caused by O3 over the continuous tri-state area of Illinois,
Iowa, and Wisconsin. The results showed that O3 concentrations reduced soybean
yield, which correlates well with the previous results from FACE- and OTC-type
experiments (Section 9.4.4.1).
Evidence is sufficient to conclude that there is a causal relationship between O3
exposure and reduced yield and quality of agricultural crops.
2.6.3 Water Cycling
Ozone can affect water use in plants and ecosystems through several mechanisms
including damage to stomatal functioning and loss of leaf area. Section 9.3.6
2-40
-------
reviewed possible mechanisms for O3 exposure effects on stomatal functioning.
Regardless of the mechanism, O3 exposure has been shown to alter stomatal
performance, which may affect plant and stand transpiration and therefore possibly
affecting hydrological cycling.
Although the evidence was from a limited number of field and modeling studies,
these findings showed an association of O3 exposure and the alteration of water use
and cycling in vegetation and ecosystems (Section 9.4.5). There is not a clear
consensus on the nature of leaf-level stomatal conductance response to O3 exposure.
When measured at steady-state high light conditions, leaf-level stomatal conductance
is often found to be reduced when exposed to O3. However, measurements of
stomatal conductance under dynamic light and vapor pressure deficit conditions
indicate sluggish responses under elevated O3 exposure which could potentially lead
to increased water loss from vegetation. In situations where stomata fail to close
under low light or water stressed conditions water loss may be greater over time.
In other situations it is possible that sluggish stomata may fail to completely open in
response to environmental stimuli and result in decreased water loss. Field studies
suggested that peak hourly O3 exposure increased the rate of water loss from several
tree species, and led to a reduction in the late-season modeled stream flow in three
forested watersheds in eastern Tennessee. Sluggish stomatal responses during O3
exposure was suggested as a possible mechanism for increased water loss during
peak O3 exposure. Currently, the O3-induced reduction in stomatal aperture is the
biological assumption for most process-based models. Therefore, results of those
models normally found that O3 reduced water loss. For example, one study found
that O3 damage and N limitation together reduced evapotranspiration and increased
runoff.
Although the direction of the response differed among studies, the evidence is
sufficient to conclude that there is likely to be a causal relationship between O3
exposure and the alteration of ecosystem water cycling.
2.6.4 Below-ground Processes
Below-ground processes are tightly linked with aboveground processes.
The responses of aboveground process to O3 exposure, such as reduced
photosynthetic rates, increased metabolic cost, and reduced root C allocation, have
provided biologically plausible mechanisms for the alteration of below-ground
processes. Since the 2006 O3 AQCD, more evidence has shown that although the
responses are often species specific, O3 altered the quality and quantity of C input to
soil, microbial community composition, and C and nutrient cycling.
Results from Aspen FACE and other experimental studies consistently found that O3
reduced litter production and altered C chemistry, such as soluble sugars, soluble
phenolics, condensed tannins, lignin, and macro/micro nutrient concentration in litter
(Section 9.4.6.1). Under elevated O3, the changes in substrate quality and quantity
could alter microbial metabolism, and therefore soil C and nutrient cycling. Several
2-41
-------
studies indicated that O3 generally suppressed soil enzyme activities
(Section 9.4.6.2). However, the impact of O3 on litter decomposition was
inconsistent and varied among species, sites, and exposure length. Similarly, O3 had
inconsistent impacts on dynamics of micro and macro nutrients (Section 9.4.6.4).
Studies from the Aspen FACE experiment suggested that the response of below-
ground C cycle to O3 exposure, such as litter decomposition, soil respiration, and soil
C content, changed over time. For example, in the early part of the experiment
(1998-2003), O3 had no impact on soil respiration but reduced the formation rates of
total soil C under elevated CO2. However, after 10 to 11 years of exposure, O3 was
found to increase soil respiration but have no substantial impact on soil C formation
under elevated CO2 (Section 9.4.6.3).
The evidence is sufficient to infer that there is a causal relationship between O3
exposure and the alteration of below-ground biogeochemical cycles.
2.6.5 Community Composition
In the 2006 O3 AQCD, the impact of O3 exposure on species competition and
community composition was assessed. Ozone was found to be one of the dominant
factors causing a decline in ponderosa and Jeffrey pine in the San Bernardino
Mountains in southern California. Ozone exposure also tended to shift the grass-
legume mixtures in favor of grass species. Since the 2006 O3 AQCD, more evidence
has shown that O3 exposure changed the competitive interactions and led to loss of
O3 sensitive species or genotypes. Studies found that the severity of O3 damage on
growth, reproduction and foliar injury varied among species (Section 9.4.3). which
provided the biological plausibility for the alteration of community composition.
Additionally, research since the last review has shown that O3 can alter community
composition and diversity of soil microbial communities.
The decline of conifer forests under O3 exposure was continually observed in several
regions. Ozone damage was believed to be an important causal factor in the dramatic
decline of sacred fir in the valley of Mexico, as well as cembran pine in southern
France and the Carpathian Mountains, although several factors, such as drought,
insect outbreak and forest management, may also contribute to or even be the
dominant factors causing the mortality of the conifer trees. Results from the Aspen
FACE site indicated that O3 could alter community composition of broadleaf forests
as well. At the Aspen FACE site, O3 reduced growth and increased mortality of a
sensitive aspen clone, while the O3 tolerant clone emerged as the dominant clone in
the pure aspen community. In the mixed aspen-birch and aspen-maple communities,
O3 reduced the competitive capacity of aspen compared to birch and maple
(Section 9.4.7.1).
The tendency for O3-exposure to shift the biomass of grass-legume mixtures in favor
of grass species was reported in the 2006 O3 AQCD and has been generally
confirmed by recent studies. However, in a high elevation mature/species-rich grass-
2-42
-------
legume pasture, O3 fumigation showed no substantial impact on community
r.nrrmnsitinn fSer.tinn Q 4 7 7s!
composition (Section 9.4.7.2)
Ozone exposure not only altered community composition of plant species, but also
microorganisms. The shift in community composition of bacteria and fungi has been
observed in both natural and agricultural ecosystems, although no general patterns
could be identified (Section 9.4.7.3).
The evidence is sufficient to conclude that there is likely to be a causal
relationship between O3 exposure and the alteration of community
composition of some ecosystems.
2.6.6 Policy Relevant Considerations
2.6.6.1 Air Quality Indices
Exposure indices are metrics that quantify exposure as it relates to measured plant
response (e.g., reduced growth). They are summary measures of monitored ambient
O3 concentrations over time intended to provide a consistent metric for reviewing
and comparing exposure-response effects obtained from various studies. No recent
information is available since 2006 that alters the basic conclusions put forth in the
2006 and 1996 O3 AQCDs. These AQCDs focused on the research used to develop
various exposure indices to help quantify effects on growth and yield in crops,
perennials, and trees (primarily seedlings). The performance of indices was
compared through regression analyses of earlier studies designed to support the
estimation of predictive O3 exposure-response models for growth and/or yield of
crops and tree (seedling) species.
Another approach for improving risk assessment of vegetation response to ambient
O3 is based on determining the O3 concentration from the atmosphere that enters the
leaf (i.e., flux or deposition). Interest has been increasing in recent years, particularly
in Europe, in using mathematically tractable flux models for O3 assessments at the
regional, national, and European scale. While some efforts have been made in the
U.S. to calculate O3 flux into leaves and canopies, little information has been
published relating these fluxes to effects on vegetation. There is also concern that not
all O3 stomatal uptake results in a yield reduction, which depends to some degree on
the amount of internal detoxification occurring with each particular species. Species
having high detoxification capacity may show little relationship between O3 stomatal
uptake and plant response. The lack of data in the U.S. and the lack of understanding
of detoxification processes have made this technique less viable for vulnerability and
risk assessments in the U.S.
2-43
-------
The main conclusions from the 1996 and 2006 O3 AQCDs regarding indices based
on ambient exposure remain valid. These key conclusions can be restated as follows:
• ozone effects in plants are cumulative;
• higher O3 concentrations appear to be more important than lower
concentrations in eliciting a response;
• plant sensitivity to O3 varies with time of day and plant development stage;
• quantifying exposure with indices that cumulate hourly O3 concentrations and
preferentially weight the higher concentrations improves the explanatory
power of exposure/response models for growth and yield, over using indices
based on mean and peak exposure values.
Various weighting functions have been used, including threshold-weighted
(e.g., SUM06) and continuous sigmoid-weighted (e.g., W126) functions. Based on
statistical goodness-of-fit tests, these cumulative, concentration-weighted indices
could not be differentiated from one another using data from previous exposure
studies. Additional statistical forms for O3 exposure indices are summarized in
Section 9.5 of this ISA. The majority of studies published since the 2006 O3 AQCD
do not change earlier conclusions, including the importance of peak concentrations,
and the duration and occurrence of O3 exposures in altering plant growth and yield.
Given the current state of knowledge and the best available data, exposure indices
that cumulate and differentially weight the higher hourly average concentrations and
also include the mid-level values continue to offer the most scientifically defensible
approach for use in developing response functions and comparing studies, as well as
for defining future indices for vegetation protection.
2.6.6.2 Exposure-Response
None of the information on effects of O3 on vegetation published since the 2006 O3
AQCD has modified the assessment of quantitative exposure-response relationships
that was presented in that document (U.S. EPA. 2006b). This assessment updates the
2006 exposure-response models by computing them using the W126 metric,
cumulated over 90 days. Almost all of the experimental research on the effects of O3
on growth or yield of plants published since 2006 used only two levels of exposure.
In addition, hourly O3 concentration data that would allow calculations of exposure
using the W126 metric are generally unavailable. However, two long-term
experiments, one with a crop species (soybean), one with a tree species (aspen), have
produced data that are used in Section 9.6 to validate the exposure-response models
presented in the 2006 O3 AQCD, and the methodology used to derive them. EPA
compared predictions from the models presented in the 2006 O3 AQCD, updated to
use the 90 day 12hr W126 metric, with more recent observations for yield of soybean
and biomass growth of trembling aspen. The models were parameterized using data
from the National Crop Loss Assessment Network (NCLAN) and EPA's National
Health and Environmental Effects Research Laboratory - Western Ecology Division
2-44
-------
(NHEERL-WED) projects, which were conducted in OTCs. The more recent
observations were from experiments using FACE technology, which is intended to
provide conditions closer to natural environments than OTC. Observations from
these new experiments were exceptionally close to predictions from the models.
The accuracy of model predictions for two widely different plant species, grown
under very different conditions, provides support for the validity of the models for
crops and trees developed using the same methodology and data for other species.
However, variability observed among species in the NCLAN and NHEERL-WED
projects indicates that the range of sensitivity between and among species is likely
quite wide.
Results from several meta-analyses have provided approximate values for responses
of yield of soybean, wheat, rice and other crops under broad categories of exposure,
relative to char coal-filtered air. Additional reports have summarized yield data for six
crop species under various broad comparative exposure categories, and reviewed 263
studies that reported effects on tree biomass. However, these analyses have proved
difficult to compare with exposure-response models, especially given that exposure
was not expressed using a common metric (i.e., W126).
2.7 The Role of Tropospheric O3 in Climate Change and UV-B
Shielding Effects
Atmospheric O3 plays an important role in the Earth's energy budget by interacting
with incoming solar radiation and outgoing infrared radiation. Tropospheric O3
makes up only a small portion of the total column of O3, but it has important
incremental effects on the overall radiation budget. Chapter 10 assesses the specific
role of tropospheric O3 in the earth's radiation budget and how perturbations in
tropospheric O3 might affect (1) climate through its role as a greenhouse gas, and
(2) health, ecology and welfare through its role in shielding the earth's surface from
solar ultraviolet (UV) radiation.
2.7.1 Tropospheric Ozone as a Greenhouse Gas
Ozone is an important greenhouse gas, and increases in its abundance in the
troposphere may contribute to climate change according to the 2007 climate
assessment by the Intergovernmental Panel on Climate Change (IPCC). Models
calculate that the global burden of tropospheric O3 has doubled since the pre-
industrial era, while observations indicate that in some regions O3 may have
increased by factors as great as 4 or 5. These increases are tied to the rise in
emissions of O3 precursors from human activity, mainly fossil fuel consumption and
agricultural processes.
Figure 2-3 shows the main steps involved in the influence of tropospheric O3 on
climate. Emissions of O3 precursors including CO, VOCs, CH4, and NOX lead to
2-45
-------
production of tropospheric O3. A change in the abundance of tropospheric O3
perturbs the radiative balance of the atmosphere, an effect quantified by the radiative
forcing metric. The earth-atmosphere-ocean system responds to the forcing with a
climate response, typically expressed as a change in surface temperature. Finally, the
climate response causes downstream climate-related health and ecosystem impacts,
such as redistribution of diseases or ecosystem characteristics due to temperature
changes. Feedbacks from both the climate response and downstream impacts can, in
turn, affect the abundance of tropospheric O3 and O3 precursors through multiple
feedback mechanisms as indicated in Figure 2-3. Direct feedbacks are discussed in
Section 10.3.2.4 and Section 10.3.3.4. while downstream climate impacts and their
feedbacks are extremely complex and outside the scope of this assessment.
Precursor Emissions of
CO, VOCs, CH4, NOX
(Tg/y)
Tropospheric
Abundance
(Tg)
Radiative Forcing
Due to O, Change
(W/m2)
0>
U.
.i
u
Climate Response
< Innate Impact1-
i on I liniinn 1 lc;t*J: t ~™*
1 mid 1 ci>N\ iicn11 'i
Note: Units shown are those typical for each quantity illustrated. Feedbacks from both the climate response and climate impacts
can, in turn, affect the abundance of tropospheric O3 and O3 precursors through multiple feedback mechanisms. Climate impacts
are deemphasized in the figure since these downstream effects are extremely complex and outside the scope of this assessment.
Figure 2-3 Schematic illustrating the effects of tropospheric O3 on climate;
including the relationship between precursor emissions,
tropospheric O3 abundance, radiative forcing, climate response,
and climate impacts.
2-46
-------
Radiative forcing by a greenhouse gas or aerosol is a metric used to quantify the
change in balance between radiation coming into and going out of the atmosphere
caused by the presence of that substance. Tropospheric O3 is a major greenhouse gas
and radiative forcing agent; evidence from satellite data shows a sharp dip in the
outgoing infrared radiation in the 9.6 |j,m O3 absorption band. Models calculate that
the global average concentration of tropospheric O3 has doubled since the
pre-industrial era, while observations indicate that in some regions O3 may have
increased by factors as great as 4 or 5. These increases are tied to the rise in
emissions of O3 precursors from human activity, mainly fossil fuel consumption and
agricultural processes. Overall, the evidence supports a causal relationship
between changes in tropospheric O3 concentrations and radiative forcing.
The impact of the tropospheric O3 change since pre-industrial times on climate has
been estimated to be about 25-40% of the anthropogenic CO2 impact and about 75%
of the anthropogenic CH4 impact according to the IPCC, ranking it third in
importance after CO2 and CH4 according to the Intergovernmental Panel on Climate
Change (IPCC) (see Section 10.3). There are large uncertainties in the magnitude of
the radiative forcing estimate attributed to tropospheric O3, making the impact of
tropospheric O3 on climate more uncertain than the effect of the longer-lived
greenhouse gases. Furthermore, radiative forcing does not take into account the
climate feedbacks that could amplify or dampen the actual surface temperature
response. Quantifying the change in surface temperature requires a complex climate
simulation in which all important feedbacks and interactions are accounted for.
The modeled surface temperature response to a given radiative forcing is highly
uncertain and can vary greatly among models and from region to region within the
same model. Even with these these uncertainties, global climate models indicate that
tropospheric O3 has contributed to observed changes in global mean and regional
surface temperatures. As a result of such evidence presented in climate modeling
studies, there is likely to be a causal relationship between changes in
tropospheric O3 concentrations and effects on climate.
2.7.2 Tropospheric Ozone and UV-B Shielding Effects
UV radiation emitted from the Sun contains sufficient energy when it reaches the
Earth to break (photolyze) chemical bonds in molecules, thereby leading to damaging
effects on living organisms and materials. Atmospheric O3 plays a crucial role in
reducing exposure to solar UV radiation at the Earth's surface. Ozone in the
stratosphere is responsible for the majority of this shielding effect, as approximately
90% of total atmospheric O3 is located there over mid-latitudes. Ozone in the
troposphere provides supplemental shielding of radiation in the wavelength band
from 280-315 nm, referred to as UV-B radiation. UV-B radiation has important
effects on human health and ecosystems, and is associated with materials damage.
Human health effects associated with solar UV-B radiation exposure include
erythema, skin cancer, ocular damage, and immune system suppression. A potential
2-47
-------
human health benefit of increased UV-B exposure involves the UV-induced
production of vitamin D which may help reduce the risk of metabolic bone disease,
type I diabetes, mellitus, and rheumatoid arthritis, and may provide beneficial
immunomodulatory effects on multiple sclerosis, insulin-dependent diabetes
mellitus, and rheumatoid arthritis. Ecosystem and materials damage effects
associated with solar UV-B radiation exposure include terrestrial and aquatic
ecosystem impacts, alteration of biogeochemical cycles, and degradation of man-
made materials.
EPA has found no published studies that adequately examine the incremental health
or welfare effects (adverse or beneficial) attributable specifically to changes in UV-B
exposure resulting from perturbations in tropospheric O3 concentrations. While the
effects are expected to be small, they cannot yet be critically assessed within
reasonable uncertainty. Overall, the evidence is inadequate to determine if a
causal relationship exists between changes in tropospheric O3 concentrations
and effects on health and welfare related to UV-B shielding.
2.8 Summary of Causal Determinations for Health Effects and
Welfare Effects
This chapter has provided an overview of the underlying evidence used in making
the causal determinations for the health and welfare effects of O3. This review builds
upon the conclusions of the previous AQCDs for O3.
The evaluation of the epidemiologic, toxicological, and controlled human exposure
studies published since the completion of the 2006 O3 AQCD have provided
additional evidence for O3-related health outcomes. Table 2-3provides an overview
of the causal determinations for all of the health outcomes evaluated. Causal
determinations for O3 and welfare effects are included in Table 2-4, while causal
determinations for climate change and UV-B shielding effects are in Table 2-5.
Detailed discussions of the scientific evidence and rationale for these causal
determinations are provided in subsequent chapters of this ISA.
2-48
-------
Table 2-3 Summary of O3 causal determinations by exposure duration and
health outcome.
Health Outcome
Conclusions from 2006 O3 AQCD
Conclusions from
this ISA
Short-Term Exposure to
Respiratory effects
The overall evidence supports a causal relationship between acute ambient
O3 exposures and increased respiratory morbidity outcomes.
Causal Relationship
Cardiovascular effects The limited evidence is highly suggestive that O3 directly and/or indirectly
contributes to cardiovascular-related morbidity, but much remains to be
done to more fully substantiate the association.
Central nervous Toxicological studies report that acute exposures to O3 are associated with
system effects alterations in neurotransmitters, motor activity, short and long term memory,
sleep patterns, and histological signs of neurodegeneration.
Total Mortality The evidence is highly suggestive that O3 directly or indirectly contributes to
non-accidental and cardiopulmonary-related mortality.
Likely to be a
Causal Relationship
Suggestive of a
Causal Relationship
Likely to be a
Causal Relationship
Long-term Exposure to O3
Respiratory effects
Cardiovascular
Effects
Reproductive and
developmental effects
Central nervous
system effects
Cancer
Total Mortality
The current evidence is suggestive but inconclusive for respiratory health
effects from long-term O3 exposure.
No studies from previous review
Limited evidence fora relationship between air pollution and birth-related
health outcomes, including mortality, premature births, low birth weights,
and birth defects, with little evidence being found for O3 effects.
Evidence regarding chronic exposure and neurobehavioral effects was not
available.
Little evidence for a relationship between chronic O3 exposure and
increased risk of lung cancer.
There is little evidence to suggest a causal relationship between chronic O3
exposure and increased risk for mortality in humans.
Likely to be a
Causal Relationship
Suggestive of a
Causal Relationship
Suggestive of a
Causal Relationship
Suggestive of a
Causal Relationship
Inadequate to infer a
Causal Relationship
Suggestive of a
Causal Relationship
2-49
-------
Table 2-4 Summary of O3 causal determination for welfare effects.
Vegetation and
Ecosystem Effects
Conclusions from 2006 O3 AQCD
Conclusions from this ISA
Visible Foliar Injury
Effects on Vegetation
Data published since the 1996 O3 AQCD strengthen
previous conclusions that there is strong evidence that
current ambient O3 concentrations cause impaired
aesthetic quality of many native plants and trees by
increasing foliar injury.
Causal Relationship
Reduced Vegetation
Growth
Data published since the 1996 O3 AQCD strengthen
previous conclusions that there is strong evidence that
current ambient O^ concentrations cause decreased
growth and biomass accumulation in annual, perennial
and woody plants, including agronomic crops, annuals,
shrubs, grasses, and trees.
Causal Relationship
Reduced Productivity in
Terrestrial Ecosystems
There is evidence that O3 is an important stressor of
ecosystems and that the effects of Os on individual plants
and processes are scaled up through the ecosystem,
affecting net primary productivity.
Causal Relationship
Reduced Carbon (C)
Sequestration in
Terrestrial Ecosystems
Limited studies from previous review
Likely to be a
Causal Relationship
Reduced Yield and
Quality of Agricultural
Crops
Data published since the 1996 O3 AQCD strengthen
previous conclusions that there is strong evidence that
current ambient O^ concentrations cause decreased yield
and/or nutritive quality in a large number of agronomic
and forage crops.
Causal Relationship
Alteration of Terrestrial
Ecosystem Water
Cycling
Ecosystem water quantity may be affected by 03
exposure at the landscape level.
Likely to be a
Causal Relationship
Alteration of Below-
ground Biogeochemical
Cycles
Ozone-sensitive species have well known responses to
Os exposure, including altered C allocation to below-
ground tissues, and altered rates of leaf and root
production, turnover, and decomposition. These shifts can
affect overall C and N loss from the ecosystem in terms of
respired C, and leached aqueous dissolved organic and
inorganic C and N.
Causal Relationship
Alteration of Terrestrial
Community Composition
Ozone may be affecting above- and below -ground
community composition through impacts on both growth
and reproduction. Significant changes in plant community
composition resulting directly from O3 exposure have
been demonstrated.
Likely to be a
Causal Relationship
2-50
-------
Table 2-5 Summary of O3 causal determination for climate and
UV-B shielding effects.
Effects
Conclusions from 2006 O3 AQCD
Conclusions from this ISA
Radiative Forcing
Climate forcing by O3 at the regional scale may be its most
important impact on climate.
Causal Relationship
While more certain estimates of the overall importance of
global-scale forcing due to tropospheric O3 await further
advances in monitoring and chemical transport modeling,
Climate Change the overall body of scientific evidence suggests that high
concentrations of O3 on the regional scale could have a
discernible influence on climate, leading to surface
temperature and hydrological cycle changes.
Likely to be a
Causal Relationship
Health and Welfare
Effects Related to
UV-B Shielding
UV-B has not been studied in sufficient detail to allow for a
credible health benefits assessment. In conclusion, the
effect of changes in surface-level O3 concentrations on
UV-induced health outcomes cannot yet be critically
assessed within reasonable uncertainty.
Inadequate to Determine if a
Causal Relationship Exists
2-51
-------
References
CAA. Clean Air Act, as amended by Pub. L. No. 101-549. section 108: Air quality criteria and control
techniques. § 7408 (1990a). http://www.law.Cornell.edu/uscode/text/42/7408
Langstaff. J. (2003). Percentiles of 1996-2000 ozone concentrations [memorandum to Joe Pinto]. Available
online
Murphy. SL; Xu. JQ; Kochanek. KD. (2012). Deaths: Preliminary data for 2010. In National Vital Statistics
Reports. (4). Hyattsville, MD: National Center for Health Statistics.
http://www.cdc.gov/nchs/data/nvsr/nvsr60/nvsr60 04.pdf
U.S. EPA (U.S. Environmental Protection Agency). (1978a). Air quality criteria for ozone and other
photochemical oxidants [EPA Report]. (EPA/600/8-78/004). Washington, DC.
U.S. EPA (U.S. Environmental Protection Agency). (1984). Air quality criteria for ozone and other
photochemical oxidants, volume III of V (review draft) [EPA Report]. (EPA-600/8-84-020A3). Research
Triangle Park, NC. http://www.ntis.gov/search/product.aspx?ABBR=PB85126050
U.S. EPA (U.S. Environmental Protection Agency). (1996a). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/P-93/004AF). Research Triangle Park,NC.
U.S. EPA (U.S. Environmental Protection Agency). (1996b). Air quality criteria for ozone and related
photochemical oxidants, Vol. II of III [EPA Report]. (EPA/600/P-93/004BF). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (2006b). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/R-05/004AF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
U.S. EPA (U.S. Environmental Protection Agency). (2008g). Notice of workshop and call for information on
integrated science assessment for ozone. Fed Reg 73: 56581-56583.
U.S. EPA (U.S. Environmental Protection Agency). (2009c). Integrated review plan for the ozone National
Ambient Air Quality Standards review (external review draft) [EPA Report]. (EPA452/D-09-001).
Washington, DC. http://www.epa. gov/ttnnaaqs/standards/ozone/data/externalreviewdraftO3IRP093 009.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2011d). Integrated review plan for the ozone National
Ambient Air Quality Standards [EPAReport]. (EPA452/R-11-006). Washington, DC.
http://www.epa.gov/ttn/naaqs/standards/ozone/data/2011 04 OzoneIRP.pdf
Zhang. L; Jacob. DJ: Downey. NV; Wood. DA; Blewitt. D; Carouge. CC: Van donkelaar. A; Jones. DBA;
Murray. IT; Wang. Y. (2011). Improved estimate of the policy-relevant background ozone in the United
States using the GEOS-Chem global model with 1/2 2/3 horizontal resolution over North America. Atmos
Environ 45: 6769-6776. http://dx.doi.Org/10.1016/i.atmosenv.2011.07.054
2-52
-------
3 ATMOSPHERIC CHEMISTRY AND AMBIENT
CONCENTRATIONS
3.1 Introduction
In the stratosphere, ozone (O3) serves the beneficial role of absorbing the Sun's
harmful ultraviolet radiation and preventing the majority of this radiation from
reaching the Earth's surface. In the troposphere, however, O3 and other
photochemical oxidants are air pollutants that can exert harmful effects on humans,
animals, and vegetation. This chapter discusses the atmospheric chemistry associated
with tropospheric O3 and other related photochemical oxidants and provides a
detailed description of their surface-level concentrations. The focus of this chapter is
on O3 since it is the NAAQS indicator for all photochemical oxidants. To the extent
possible, other photochemical oxidants are discussed, but limited information is
currently available. Although O3 is involved in reactions in indoor air, the focus in
this chapter will be on chemistry occurring in outdoor, ambient air.
The material in this chapter is organized as follows. Section 3.2 outlines the physical
and chemical processes involved in O3 formation and removal. Section 3.3 describes
the latest methods used to model global O3 concentrations, and Section 3.4 describes
the application of these methods for estimating background concentrations of O3 that
are useful for risk and policy assessments informing decisions about the NAAQS.
Section 3.5 includes a comprehensive description of available O3 monitoring
techniques and monitoring networks, while Section 3.6 presents information on the
spatial and temporal variability of O3 concentrations across the U.S. and their
associations with other pollutants using available monitoring data. Section 3.7
summarizes the main conclusions from Chapter 3. Finally, Section 3.8 provides
supplemental material on atmospheric model simulations of background O3
concentrations (referenced in Section 3.4) and Section 3.9 contains supplemental
material on observed ambient O3 concentrations (referenced in Section 3.6).
3.2 Physical and Chemical Processes
Ozone in the troposphere is a secondary pollutant formed by photochemical reactions
of precursor gases and is not directly emitted from specific sources. Ozone and other
oxidants, such as peroxyacetyl nitrate (PAN) and H2O2 form in polluted areas by
atmospheric reactions involving two main classes of precursor pollutants: VOCs and
NOx.1 Carbon monoxide (CO) is also important for O3 formation in polluted areas
and in the remote troposphere. The formation of O3, other oxidants and oxidation
products from these precursors is a complex, nonlinear function of many factors
1 The term VOCs refers to all organic gas-phase compounds in the atmosphere, both biogenic and anthropogenic in origin. This
definition excludes CO and CO2. NOX, also referred to as nitrogen oxides, is equal to the sum of NO and NO2.
3-1
-------
including (1) the intensity and spectral distribution of sunlight; (2) atmospheric
mixing; (3) concentrations of precursors in the ambient air and the rates of chemical
reactions of these precursors; and (4) processing on cloud and aerosol particles.
Ozone is present not only in polluted urban atmospheres, but throughout the
troposphere, even in remote areas of the globe. The same basic processes involving
sunlight-driven reactions of NOX, VOCs, and CO contribute to O3 formation
throughout the troposphere. These processes also lead to the formation of other
photochemical products, such as PAN, HNO3, and H2SO4, and to other compounds,
such as HCHO and other carbonyl compounds, and to secondary components of
particulate matter.
A schematic overview of the major photochemical cycles influencing O3 in the
troposphere and the stratosphere is given in Figure 3-1. Included in the figure are
reactions involving radicals derived from man-made chemicals and that are
responsible for depleting stratospheric O3. Most (approximately 90%) of the total O3
column in the earth's atmosphere resides in the stratosphere, and it is responsible for
absorbing harmful solar ultraviolet radiation, the harmful effects of which are
discussed in Chapter 10. This solar ultraviolet radiation also initiates the
photochemical reactions that are responsible for producing O3 in the troposphere.
The processes responsible for producing summertime O3 episodes are fairly well
understood, and were covered in detail in the 2006 O3 AQCD (U.S. EPA. 2006b).
This section focuses on topics that form the basis for discussions in later chapters,
and for which there is substantial new information since the previous O3 review.
3-2
-------
Figure 3-1 Schematic overview of photochemical processes influencing
stratospheric and tropospheric Os.
Major episodes of high O3 concentrations in the eastern U.S. and in Europe are
associated with slow moving high pressure systems. High pressure systems during
the warmer seasons are associated with the sinking of air, resulting in warm,
generally cloudless skies, with light winds. The sinking of air results in the
development of stable conditions near the surface which inhibit or reduce the vertical
mixing of O3 precursors. Photochemical activity involving these precursors is
enhanced because of higher temperatures and the availability of sunlight during the
warmer seasons. In the eastern U.S., concentrations of O3 and other secondary
pollutants are determined by meteorological and chemical processes extending
typically over areas of several hundred thousand square kilometers (Civerolo et al.,
2003; Rao et al., 2003). Ozone episodes are thus best regarded as regional in nature.
The conditions conducive to formation of high O3 can persist for several days. These
conditions have been described in greater detail in the 1996 and 2006 O3 AQCDs
(U.S. EPA, 2006b, 1996a). The transport of pollutants downwind of major urban
centers is characterized by the development of urban plumes. Mountain barriers limit
mixing (as in Los Angeles and Mexico City) and result in a higher frequency and
duration of days with high O3 concentrations. However, orographic lifting over the
San Gabriel Mountains results in O3 transport from Los Angeles to areas hundreds of
kilometers downwind (e.g., in Colorado and Utah) (Langford et al., 2009). Ozone
concentrations in southern urban areas (such as Houston, TX and Atlanta, GA) tend
3-3
-------
to decrease with increasing wind speed. In northern U.S. cities (such as Chicago, IL;
New York, NY; Boston, MA; and Portland, ME), the average O3 concentrations over
the metropolitan areas increase with wind speed, indicating that transport of O3 and
its precursors from upwind areas is important (Schichtel and Husar, 2001; Husar and
Renard. 1998).
Aircraft observations indicate that there can be substantial differences in mixing
ratios of key species between the surface and the overlying atmosphere (Berkowitz
and Shaw, 1997; Fehsenfeld et al., 1996). In particular, mixing ratios of O3 can
(depending on time and location) be higher in the lower free troposphere (aloft) than
in the planetary boundary layer (PBL) during multiday O3 episodes (Taubman et al.,
2006; Taubman et al., 2004). Convective processes and turbulence transport O3 and
other pollutants both upward and downward throughout the planetary boundary layer
and the free troposphere. During the day, convection (driven by heating of the earth's
surface) results in a deeper PBL, with vertically well mixed O3 and precursors.
As solar heating of the surface decreases going into night, the daytime boundary
layer collapses leaving behind O3 and its precursors in a residual layer above a
shallow nighttime boundary layer. Pollutants in the residual layer have now become
essentially part of the free troposphere, as described in Annex AX2.3.2 of the 2006
O3 AQCD (U.S. EPA, 2006b). Winds in the free troposphere tend to be stronger than
those closer to the surface and so are capable of transporting pollutants over long
distances. Thus, O3 and its precursors can be transported vertically by convection
into the upper part of the mixed layer on one day, then transported overnight as a
layer of elevated mixing ratios, and then entrained into a growing convective
boundary layer downwind and brought back down to the surface.
High O3 concentrations showing large diurnal variations at the surface in southern
New England were associated with the presence of such layers (Berkowitz et al.,
1998). Winds several hundred meters above the ground can bring pollutants from the
west, even though surface winds are from the southwest during periods of high O3 in
the eastern U.S. (Blumenthal et al., 1997). These considerations suggest that in many
areas of the U.S., O3 and its precursors can be transported over hundreds if not
thousands of kilometers.
Nocturnal low level jets (LLJs) are an efficient means for transporting pollutants that
have been entrained into the residual boundary layer over hundreds of kilometers.
LLJs are most prevalent in the central U.S. extending northward from eastern Texas,
and along the Atlantic states extending southwest to northeast. LLJs have also been
observed off the coast of California. Turbulence induced by wind shear associated
with LLJs brings pollutants to the surface and results in secondary O3 maxima during
the night and early morning in many locations (Corsmeier et al.. 1997). Comparison
of observations at low elevation surface sites with those at nearby high elevation sites
at night can be used to discern the effects of LLJs. For example, Fischer (2004)
found occasions when O3 at the base of Mt. Washington during the night was much
higher than typically observed, and closer to those observed at the summit of Mt.
Washington. They suggested that mechanically driven turbulence due to wind shear
caused O3 from aloft to penetrate the stable nocturnal inversion thus causing O3 to
3-4
-------
increase near the base of Mt. Washington. The high wind speeds causing this
mechanically driven turbulence could have resulted from the development of a LLJ.
Stratospheric intrusions and intercontinental transport of O3 are also important and
are covered in Section 3.4 in relation to background concentrations.
3.2.1 Sources of Precursors Involved in Os Formation
Emissions of O3 precursor compounds (NOX, VOCs, and CO) can be divided into
natural and anthropogenic source categories. Natural sources can be further divided
into biogenic from vegetation, microbes, and animals, and abiotic from biomass
combustion, lightning, and geogenic sources. However, the distinction between
natural and anthropogenic sources is often difficult to make in practice, as human
activities directly or indirectly affect emissions from what would have been
considered natural sources during the pre-industrial era. Thus, emissions from plants
and animals used in agriculture have been referred to as anthropogenic or biogenic in
different applications. Wildfire emissions can be considered natural, except that
forest management practices can lead to buildup of fuels on the forest floor, thereby
altering the frequency and severity of forest fires.
Estimates of emissions for NOX, VOCs, and CO from the 2005 National Emissions
Inventory (NEI) (U.S. EPA. 2008a) are shown in Figure 3-2 to provide a general
indication of the relative importance of the different sources in the U.S. as a whole.
The magnitudes of the sources are strongly location and time dependent and so
should not be used to apportion sources of exposure. Shown in Figure 3-2 are Tier 1
categories. The miscellaneous category can be quite large compared to total
emissions, especially for CO and VOCs. The miscellaneous category includes
agriculture and forestry, wildfires, prescribed burns, and a much more modest
contribution from structural fires.
3-5
-------
Nitrogen Oxides (NOJ
Tolal Emissions = 1 7.4 MT
HIGHWAY VEHICLES I
OFF-HIGHWAY I
ISQ
144
FUEL COMB. ELEC. UTIL. I
134
FUEL COMB. INDUSTRIAL ••••••••
FUEL COMB. OTHER BL I 0 66
OTHER INDUSTRIAL PROCESSES I - 1 0.44
PETROLEUM & RELATED INDUSTRIES BO. 32
MISCELLANEOUS O0.25
WASTE DISPOSAL & RECYCLING DO. 13
METALS PROCESSING 10.06
CHEMICALS. ALLIED PRODUCT MFG 10.05
STORAGE & TRANSPORT 10.015
SOLVENT UTILIZATION 1 0.004
1.6
Volatile Organic Compounds (VOC)
Total Emissions = 16.7 MT
HIGHWAY VEHICLES I
OFF-HIGHWAY
FUEL COMB. ELEC. UTIL. 00.04
FUEL COMB INDUSTRIAL dO 12
FUEL COMB. OTHER 1053
OTHER INDUSTRIAL PROCESSES >i <"41
PETROLEUM & RELATED INDUSTRIES I IO.S1
MISCELLANEOUS I ~
WASTE DISPOSAL & RECYCLING I inafi
METALS PROCESSING H0.04
CHEMICAL & ALLIED PRODUCT MFC CZl 0.21
STORAGE & TRANSPORT I
SOLVENT UTILIZATION
1 1 .3
Carbon Monoxide (CO)
Tolal Emissions = 84.6 MT
HIGHWAY VEHICLES
FUEL COMB ELEC. UTIL. 10 58
FUEL COMB INDUSTRIAL B 1 04
FUEL COMB. OTHER IH3.02
OTHER INDUSTRIAL PROCESSES D0.48
PETROLEUM & RELATED INDUSTRIES 1032
MISCELLANEOUS
WASTE DISPOSAL & RECYCLING • 1.41
METALS PROCESSING B0.75
CHEMICAL SALLIED PRODUCT MFC 10.19
STORAGE & TRANSPORT 10.1
SOLVENT UTILIZATION 0.002
1137
4567
Emissions (Millions Tons/Year)
~\ A 73
l?fi
I298
13.85
34
Emissions (Millions Tons/Year)
14404
10 15 20 25 30 35 40 45 50
Emissions (Millions Tons/Year)
Note: NOX (top), VOCs (middle), and CO (bottom) in the U.S. in million metric tons (MT) per year.
Source: U.S. EPA (2008a).
Figure 3-2 Estimated anthropogenic emissions of O3 precursors for 2005.
3-6
-------
Anthropogenic NOX emissions are associated with combustion processes. Most
emissions are in the form of NO, which is formed at high combustion temperatures
from atmospheric nitrogen (N2) and oxygen (O2) and from fuel nitrogen (N).
According to the 2005 NEI, the largest sources of NOX are on- and off-road (such as
construction equipment, agricultural equipment, railroad trains, ships, and aircraft)
mobile sources and electric power generation plants. Emissions of NOX therefore are
highest in areas having a high density of power plants and in urban regions having
high traffic density. Dallmann and Harlev (2010) compared NOX emissions estimates
from the 2005 NEI mobile sector data with an alternative method based on fuel
consumption and found reasonable agreement in total U.S. anthropogenic emissions
between the two techniques (to within about 5%). However, emissions from on-road
diesel engines in the fuel based inventory constituted 46% of total mobile source
NOX compared to 35% in the EPA inventory. As a result, emissions from on-road
diesel engines in the fuel based approach are even larger than electric power
generation as estimated in the 2005 NEI, and on-road diesel engines might represent
the largest single NOX source category. Differences between the two techniques are
largely accounted for by differences in emissions from on-road gasoline engines.
Uncertainties in the fuel consumption inventory ranged from 3% for on-road gasoline
engines to 20% for marine sources, and in the EPA inventory uncertainties ranged
from 16% for locomotives to 30% for off-road diesel engines. It should be noted that
the on-road diesel engine emissions estimate by Dallmann and Harley (2010) is still
within the uncertainty of the EPA estimate (22%). Because of rapid changes to heavy
duty diesel NOX controls, emissions are likely to also rapidly change.
Satellite-based techniques have been used to obtain tropospheric concentrations of
O3 precursors (e.g., NO2, VOCs, and CO). Such satellite-based measurements
provide a large-scale picture of spatial and temporal distribution of NO2, VOCs, and
CO that can be used to evaluate emissions inventories produced using the bottom-up
approach and to produce top-down emissions inventories of these species. Although
there are uncertainties associated with satellite-based measurements, several studies
have shown the utility of top-down constraints on the emissions of O3 precursors
(McDonald-Buller et al.. 2011 and references therein). Following mobile sources,
power plants are considered the second largest anthropogenic source of NOX. Over
the past decade, satellite measurements have shown appreciable reductions in NOX
power plant emissions across the U.S. as a result of emission abatement strategies
(Stavrakou et al.. 2008: Kim et al.. 2006). For instance, Kim et al. (2006) observed a
34% reduction in NOX emission over the Ohio River Valley from 1999-2006 due to
such strategies. Based on these results, less than 25% of anthropogenic NOX
emissions were expected to originate from power plants in this region. Uncertainty in
NOX satellite measurements are impacted by several factors, such as cloud and
aerosol properties, surface albedo, stratospheric NOX concentration, and solar zenith
angle. Boersma et al. (2004) estimated an overall uncertainty between 35-60% for
satellite-retrieved NOX measurements in urban, polluted regions. Although trends in
satellite-retrieved NOX power plant emissions reported by Kim et al. (2006) are
uncertain to some extent, similar reductions were reported by region-wide power
plant measurements (e.g., Continuous Emission Monitoring System observations,
CEMS).
3-7
-------
Major natural sources of NOX in the U.S. include lightning, soils, and wildfires.
Uncertainties in natural NOX emissions are much larger than for anthropogenic NOX
emissions. Fang et al. (2010) estimated lightning generated NOX of-0.6 MT for July
2004. This value is -40% of the anthropogenic emissions for the same period, but the
authors estimated that -98% is formed in the free troposphere and so contributions to
the surface NOX burden are low because most of this NOX is oxidized to nitrate
containing species during downward transport into the planetary boundary layer.
The remaining 2% is formed within the planetary boundary layer. Both nitrifying and
denitrifying organisms in the soil can produce NOX, mainly in the form of NO.
Emission rates depend mainly on fertilization amount and soil temperature and
moisture. Nationwide, about 60% of the total NOX emitted by soils is estimated to
occur in the central corn belt of the United States. Spatial and temporal variability in
soil NOX emissions leads to considerable uncertainty in emissions estimates.
However, these emissions are relatively low, only -0.97 MT/year, or about 6% of
anthropogenic NOX emissions. However, these emissions occur mainly during
summer when O3 is of most concern and occur across the entire country including
areas where anthropogenic emissions are low.
Hundreds of VOCs, containing mainly 2 to -12 carbon (C) atoms, are emitted by
evaporation and combustion processes from a large number of anthropogenic
sources. The two largest anthropogenic source categories in the U.S. EPA's
emissions inventories are industrial processes and transportation. Emissions of VOCs
from highway vehicles account for roughly two-thirds of the transportation-related
emissions. The accuracy of VOC emission estimates is difficult to determine, both
for stationary and mobile sources. Evaporative emissions, which depend on
temperature and other environmental factors, compound the difficulties of assigning
accurate emission factors. In assigning VOC emission estimates to the mobile source
category, models are used that incorporate numerous input parameters (e.g., type of
fuel used, type of emission controls, and age of vehicle), each of which has some
degree of uncertainty.
On the U.S. and global scales, emissions of VOCs from vegetation are much larger
than those from anthropogenic sources. Emissions of VOCs from anthropogenic
sources in the 2005 NEI were -17 MT/year (wildfires constitute -1/6 of that total
and were included in the 2005 NEI under the anthropogenic category, but see
Section 3.4 for how wildfires are treated for background O3 considerations), but were
29 MT/year from biogenic sources. Uncertainties in both biogenic and anthropogenic
VOC emission inventories prevent determination of the relative contributions of
these two categories, at least in many areas. Vegetation emits substantial quantities of
VOCs, such as terpenoid compounds (isoprene, 2-methyl-3-buten-2-ol,
monoterpenes), compounds in the hexanal family, alkenes, aldehydes, organic acids,
alcohols, ketones, and alkanes. The major chemicals emitted by plants are isoprene
(40%), other terpenoid and sesqui-terpenoid compounds (25%) and the remainder
consists of assorted oxygenated compounds and hydrocarbons according to the 2005
NEI. Most biogenic emissions occur during the summer because of their dependence
on temperature and incident sunlight. Biogenic emissions are also higher in southern
states than in northern states for these reasons and because of species variations.
3-8
-------
The uncertainty in natural emissions is about 50% for isoprene under midday
summer conditions and could be as much as a factor often higher for some
compounds (Guenther et al., 2000). In EPA's regional modeling efforts, biogenic
emissions of VOCs are estimated using the Biogenic Emissions Inventory System
(BEIS) model (U.S. EPA, 201 Ob) with data from the Biogenic Emissions Landuse
Database (BELD) and annual meteorological data. However, other emissions models
are used such as Model of Emissions of Gases and Aerosols from Nature (MEGAN)
(Guenther et al.. 2006). especially in global modeling efforts.
Satellite measurements of HCHO, produced by the oxidation of isoprene and other
VOCs, have also been used to estimate biogenic VOC emissions attributed to
isoprene (Millet et al., 2008; Millet et al., 2006). Millet et al. (2008) demonstrated
that both satellite-based and model techniques capture the spatial variability of
biogenic isoprene emissions in the U.S. reasonably well (satellite [versus MEGAN]
isoprene estimates, R2 = 0.48 or 0.68 depending on vegetation data base used).
However, MEGAN tends to overestimate emissions compared to satellite-based
measurements. The uncertainty in satellite derived isoprene emissions is roughly
40%, based on combined uncertainty in satellite retrieval and isoprene yield from
isoprene oxidation (Millet et al., 2006), which is similar to the error associated with
model-based techniques (-50%) (e.g., Millet et al., 2006; Guenther et al., 2000).
Anthropogenic CO is emitted primarily by incomplete combustion of carbon-
containing fuels. In general, any increase in fuel oxygen content, burn temperature,
or mixing time in the combustion zone will tend to decrease production of CO
relative to CO2. However, it should be noted that controls mute the response of CO
formation to fuel-oxygen. CO emissions from large fossil-fueled power plants are
typically very low since the boilers at these plants are tuned for highly efficient
combustion with the lowest possible fuel consumption. Additionally, the CO-to-CO2
ratio in these emissions is shifted toward CO2 by allowing time for the furnace flue
gases to mix with air and be oxidized by OH to CO2 in the hot gas stream before the
OH concentrations drop as the flue gases cool. Nationally, on-road mobile sources
constituted about half of total CO emissions in the 2005 NEI. When emissions from
off-highway (non-road) vehicles are included, it can be seen from Figure 3-2 that all
mobile sources accounted for about three-quarters of total anthropogenic CO
emissions in the U.S.
Analyses by Harlev et al. (2005) and Parrish (2006) are consistent with the
suggestion in Pollack et al. (2004) that the EPA MOBILE6 vehicle emissions model
(U.S. EPA. 2010d) overestimates vehicle CO emissions by a factor of ~2. Field
measurements by Bishop and Stedman (2008) were in accord with the Parrish (2006)
findings that the measured trends of CO and NOX concentrations from mobile
sources in the U.S. indicated that modeled CO emission estimates were substantially
too high. Hudman et al. (2008) found that the NEI overestimated anthropogenic CO
emissions by 60% for the eastern U.S. during the period July 1-August 15, 2004
based on comparison of aircraft observations of CO from the International
Consortium for Atmospheric Research on Transport and Transformation (ICARTT)
campaign (Fehsenfeld et al.. 2006) and results from a tropospheric chemistry model
3-9
-------
(GEOS-Chem). Improvements in emissions technologies not correctly represented in
MOBILE emission models have been suggested as one cause for this discrepancy.
For example, Pokharel et al. (2003, 2002) demonstrated substantial decrements in the
CO fraction of tailpipe exhaust in several U.S. cities and Burgard et al. (2006)
documented improvements in emission from heavy-duty on-road diesel engines.
Some of the largest errors in the MOBILE models are addressed in the successor
model, MOVES (U.S. EPA. 201 If).
Estimates of biogenic CO emissions in the 2005 NEI are made in a manner similar to
that for VOCs. National biogenic emissions, excluding fires, were estimated to
contribute ~7% and wildfires added another -16% to the national CO emissions total.
Photodecomposition of organic matter in oceans, rivers, lakes, and other surface
waters, and from soil surfaces also releases CO (Goldstein and Galbally, 2007).
However, soils can act as a CO source or a sink depending on soil moisture, UV flux
reaching the soil surface, and soil temperature (Conrad and Seiler, 1985). Soil uptake
of CO is driven by anaerobic bacteria (Inman et al., 1971). Emissions of CO from
soils appear to occur by abiotic processes, such as thermodecomposition or
photodecomposition of organic matter. In general, warm and moist conditions found
in most soils favor CO uptake, whereas hot and dry conditions found in deserts and
some savannas favor the release of CO (King, 1999).
3.2.2 Gas Phase Reactions Leading to O3 Formation
Photochemical processes involved in O3 formation have been extensively reviewed
in a number of books (Jacobson. 2002: Jacob. 1999: Seinfeld and Pandis. 1998:
Finlavson-Pitts and Pitts. 1986) and in the 1996 and 2006 O3 AQCDs (U.S. EPA.
2006b. 1996a). The photochemical formation of O3 in the troposphere proceeds
through the oxidation of NO to nitrogen dioxide (NO2) by organic-peroxy (RO2) or
hydro-peroxy (HO2) radicals. The peroxy radicals oxidizing NO to NO2 are formed
during the oxidation of VOCs as presented in Annex AX2.2.2 of the 2006 O3 AQCD
(U.S. EPA. 2006b). The photolysis of NO2 yields NO and a ground-state oxygen
atom, O(3P), which then reacts with molecular oxygen to form O3.
VOCs important for the photochemical formation of O3 include alkanes, alkenes,
aromatic hydrocarbons, carbonyl compounds (e.g., aldehydes and ketones), alcohols,
organic peroxides, and halogenated organic compounds (e.g., alkyl halides). This
array of compounds encompasses a wide range of chemical properties and lifetimes:
isoprene has an atmospheric lifetime of approximately an hour, whereas methane has
an atmospheric lifetime of about a decade.
In urban areas, compounds representing all classes of VOCs and CO are important
for O3 formation. In non-urban vegetated areas, biogenic VOCs emitted from
vegetation tend to be the most important. In the remote troposphere, methane (CH4)
and CO are the main carbon-containing precursors to O3 formation. The oxidation of
VOCs is initiated mainly by reaction with hydroxyl (OH) radicals. The primary
source of OH radicals in the atmosphere is the reaction of electronically excited
3-10
-------
oxygen atoms, O^D), with water vapor. O(1U) is produced by the photolysis of O3 in
the Hartley bands. In polluted areas, the photolysis of aldehydes (e.g., HCHO),
HONO and H2O2 can also be appreciable sources of OH, or HO2 radicals that can
rapidly be converted to OH (Eisele et al., 1997). Ozone can oxidize alkenes, as can
NO3 radicals. NO3 radicals are most effective at night when they are most abundant.
In coastal environments and other selected environments, atomic Cl and Br radicals
can also initiate the oxidation of VOCs as discussed in Annex AX2.2.3 of the 2006
O3 AQCD (U.S. EPA. 2006b). It was also emphasized in Annex AX2.2.9 of the 2006
O3 AQCD (U.S. EPA. 2006b) that the reactions of oxygenated VOCs are important
components of O3 formation. They may be present in ambient air not only as the
result of the atmospheric oxidation of hydrocarbons but also by direct emissions. For
example, motor vehicles (including compressed natural gas vehicles) and some
industrial processes emit formaldehyde (Rappengluck et al.. 2009) and vegetation
emits methanol.
There are a large number of oxidized N-containing compounds in the atmosphere
including NO, NO2, NO3, HNO2, HNO3, N2O5, HNO4, PAN (and its homologues),
and other organic nitrates (such as alkyl nitrates, isoprene nitrates), and particulate
nitrate. Collectively these species are referred to as NOY. Oxidized nitrogen
compounds are emitted to the atmosphere mainly as NO, which rapidly interconverts
with NO2 and so NO and NO2 are often "lumped" together into their own group or
family, which is referred to as NOX. All the other species mentioned above in the
definition of NOY are products of NOX reactions are referred to as NOZ, such that
NOY = NOX + NOZ. The major reactions involving interconversions of oxidized
N species were covered in Annex AX2.2.4 of the 2006 O3 AQCD (U.S. EPA.
2006b). Mollner et al. (2010) identified pernitrous acid (HOONO), an unstable
isomer of nitric acid, as a product of the major gas phase reaction forming HNO3.
However, since pernitrous acid is unstable, it is not a substantial reservoir for NOX.
This finding stresses the importance of identifying products in addition to measuring
the rate of disappearance of reactants in kinetic studies.
The photochemical cycles by which the oxidation of hydrocarbons leads to O3
production are best understood by considering the oxidation of methane, structurally
the simplest VOC. The CH4 oxidation cycle serves as a model for the chemistry of
the relatively clean or unpolluted troposphere (although this is a simplification
because vegetation releases large quantities of complex VOCs, such as isoprene, into
the atmosphere). In the polluted atmosphere, the underlying chemical principles are
the same, as discussed in Annex AX2.2.5 of the 2006 O3 AQCD (U.S. EPA. 2006b).
The conversion of NO to NO2 occurring with the oxidation of VOCs is accompanied
by the production of O3 and the efficient regeneration of the OH radical, which in
turn can react with other VOCs as shown in Figure 3-1.
The oxidation of alkanes and alkenes in the atmosphere has been treated in depth in
the 1996 O3 AQCD (U.S. EPA. 1996a) and was updated in Annexes AX2.2.6 and
AX2.2.7 of the 2006 O3 AQCD (U.S. EPA. 2006b). In contrast to simple
hydrocarbons containing one or two C atoms, detailed kinetic information about the
gas phase oxidation pathways of many anthropogenic hydrocarbons (e.g., aromatic
3-11
-------
compounds such as benzene and toluene), biogenic hydrocarbons (e.g., isoprene, the
monoterpenes), and their intermediate oxidation products (e.g., peroxides, nitrates,
carbonyls and epoxides) is lacking. This information is crucial even for compounds
formed in low yields, such as isoprene epoxides, as they are major precursors to
secondary organic aerosol formation (see, e.g., Surratt et al., 2010). Reaction with
OH radicals represents the major loss process for alkanes. Reaction with chlorine
(Cl) atoms is an additional sink for alkanes. Stable products of alkane photo-
oxidation are known to include a wide range of compounds and concentrations
including carbonyl compounds, alkyl nitrates, and d-hydroxycarbonyls. Major
uncertainties in the atmospheric chemistry of the alkanes concern the chemistry of
alkyl nitrate formation; these uncertainties affect the amount of NO-to-NO2
conversion occurring and, hence, the amounts of O3 formed during photochemical
degradation of the alkanes.
The reaction of OH radicals with aldehydes produced during the oxidation of alkanes
forms acyl (R'CO) radicals, and acyl peroxy radicals (R'C(O)—O2) are formed by the
further addition of O2. As an example, the oxidation of ethane (C2H5-H) yields
acetaldehyde (CH3-CHO). The reaction of CH3-CHO with OH radicals yields
acetyl radicals (CH3—CO). The acetyl radicals will then participate with O2 in a
termolecular recombination reaction to form acetyl peroxy radicals, which can then
react with NO to form CH3 + CO2 or they can react with NO2 to form PAN. PAN
acts as a temporary reservoir for NO2. Upon the thermal decomposition of PAN,
either locally or elsewhere, NO2 is released to participate in the O3 formation process
again.
Alkenes react in ambient air with OH, NO3, and Cl radicals and with O3. All of these
reactions are important atmospheric transformation processes, and all proceed by
initial addition to the carbon double bonds. Major products of alkene photo-oxidation
include carbonyl compounds. Hydroxynitrates and nitratocarbonyls, and
decomposition products from the energy-rich biradicals formed in alkene-O3
reactions are also produced. Major uncertainties in the atmospheric chemistry of the
alkenes concern the products and mechanisms of their reactions with O3, especially
the yields of radicals that participate in O3 formation. Examples of oxidation
mechanisms of complex alkanes and alkenes can be found in comprehensive texts
such as Seinfeld and Pandis (1998).
Although the photochemistry of isoprene is crucial for understanding O3 formation,
there are major uncertainties in its oxidation pathways that still need to be addressed.
Apart from the effects of the oxidation of isoprene on production of radicals and O3
formation, isoprene nitrates (RONO2) appear to play an important role as NOX
reservoirs over the eastern U.S. (e.g.. Perring et al.. 2009). Their decomposition leads
to the recycling of NOX, which can participate in the O3 formation process.
Laboratory and field-based approaches support yields for RONO2 formation from
isoprene oxidation ranging from 4% to 12% (see summaries in, Lockwood et al..
2010: Perring et al.. 2009: Horowitz et al.. 2007: von Kuhlmann et al.. 2004).
The rate at which RONO2 reacts to recycle NOX is poorly understood (Archibald et
al.. 2010: Paulot et al.. 2009) with ranges from 0 to 100% in global chemical
3-12
-------
transport models. This range affects the sign of the O3 response to changes in
biogenic VOC emissions as well as the sensitivity of O3 to changes in NOX
emissions (Archibald et al., 2011; Ito et al., 2009; Weaver et al., 2009; Horowitz et
al., 2007; Fiore et al., 2005). In models that assume zero RONO2 recycling (Zhang et
al., 2011; Wu et al., 2007; Fiore et al., 2003) O3 production is suppressed relative to
a model that recycles NOX from RONO2 (Kang et al.. 2003). A related issue
concerns the lack of regeneration of OH + HO2 radicals especially in low NOX
(<~1 ppb) environments. The isomerization of the isoprene peroxy radicals that are
formed after initial OH attack and subsequent reactions could help resolve this
problem (Peeters and Muller. 2010; Peeters et al.. 2009) and result in increases in OH
concentrations from 20 to 40% over the southeastern U.S. (Archibald et al.. 2011).
However, the effectiveness of this pathway is uncertain and depends on the fraction
of isoprene-peroxy radicals reacting by isomerization. Crounse et al. (2011)
estimated that only 8-11% of the isoprene-peroxy radicals isomerizes to reform HO2
radicals. Hofzumahaus et al. (2009) also found under predictions of OH in the Pearl
River Delta and they also note that the sequence of reactions beginning with OH
attack on VOCs introduces enormous complexity which is far from being fully
understood.
The oxidation of aromatic hydrocarbons constitutes an important component of the
chemistry of O3 formation in urban atmospheres as discussed in Annex AX2.2.8 of
the 2006 O3 AQCD (U.S. EPA. 2006b). Virtually all of the important aromatic
hydrocarbon precursors emitted in urban atmospheres are lost through reaction with
the hydroxyl radical. Loss rates for these compounds vary from slow (e.g., benzene)
to moderate (e.g., toluene), to very rapid (e.g., xylene and trimethylbenzene isomers).
However, the mechanism for the oxidation of aromatic hydrocarbons following
reaction with OH is poorly understood, as is evident from the poor mass balance of
the reaction products. The mechanism for the oxidation of toluene has been studied
most thoroughly, and there is general agreement on the initial steps in the
mechanism. However, at present there is no promising approach for resolving the
remaining issues concerning the later steps. The oxidation of aromatic hydrocarbons
also leads to particle formation that could remove gas-phase constituents that
participate in O3 formation.
Adequate analytical techniques needed to identify and quantify key intermediate
species are not available for many compounds. In addition, methods to synthesize
many of the suspected intermediate compounds are not available so that laboratory
studies of their reaction kinetics cannot be performed. Similar considerations apply to
the oxidation of biogenic hydrocarbons besides isoprene. These considerations are
important because oxidants (other than O3) that are formed from the chemistry
described above could exert effects on human health and perhaps also on vegetation
(Doyle et al., 2007; Doyle et al., 2004; Sexton et al., 2004). Gas phase oxidants
include PAN, H2O2, CH3OOH, and other organic hydroperoxides.
Ozone is lost through a number of gas phase reactions and deposition to surfaces.
The reaction of O3 with NO to produce NO2, for example in urban centers near
roads, mainly results in the recycling of O3 downwind via the recombination of
3-13
-------
O(3P) with O2 to re-form O3. By itself, this reaction does not lead to a net loss of O3
unless the NO2 is converted to stable end products such as HNO3. Ozone reacts with
unsaturated hydrocarbons and with OH and HO2 radicals.
Perhaps the most recent field study aimed at obtaining a better understanding of
atmospheric chemical processes was the Second Texas Air Quality Field Study
(TexAQS-II) conducted in Houston in August and September 2006 (Olaguer et al.,
2009). The TexAQS-II Radical and Aerosol Measurement Project (TRAMP) found
evidence for the importance of short-lived radical sources such as HCHO and HONO
in increasing O3 productivity. During TRAMP, daytime HCHO pulses as large as
32 ppb were observed and attributed to industrial activities upwind in the Houston
Ship Channel (HSC) and HCHO peaks as large as 52 ppb were detected by in situ
surface monitors in the HSC. Primary HCHO produced in flares from local refineries
and petrochemical facilities could increase peak O3 by -30 ppb (Webster et al.,
2007). Other findings from TexAQS-II included substantial concentrations of HONO
during the day, with peak concentrations approaching 1 ppb at local noon. These
concentrations are well in excess of current air quality model predictions using gas
phase mechanisms alone (Sarwar et al., 2008) and multiphase processes are needed
to account for these observations. Olaguer et al. (2009) also noted that using
measured HONO brings modeled O3 concentrations into much better agreement with
observations and could result in the production of an additional 10 ppb O3. Large
nocturnal vertical gradients indicating a surface or near-surface source of HONO,
and large concentrations of night-time radicals (~30 ppt HO2) were also found during
TRAMP.
3.2.3 Multiphase Processes
In addition to the gas phase, chemical reactions also occur on the surfaces of or
within cloud droplets and airborne particles. Their collective surface area is huge,
implying that collisions with gas phase species occur on very short time scales.
In addition to hydrometeors (e.g., cloud and fog droplets and snow and ice crystals)
there are also potential reactions involving atmospheric particles of varying
composition (e.g., wet [deliquesced] inorganic particles, mineral dust, carbon chain
agglomerates and organic carbon particles). Multiphase reactions are involved in the
formation of a number of species such as particulate nitrate, and gas phase HONO
that can act to both increase and reduce the rate of O3 formation in the polluted
troposphere. Data collected in Houston as part of TexAQS-II summarized by Olaguer
et al. (2009) indicate that concentrations of HONO are much higher than can be
explained by gas phase chemistry and by tailpipe emissions. Photolysis of HONO
formed in multiphase reactions in addition to the other sources can help to reduce the
model underestimate of simulated O3 in Houston.
Multiphase processes have been associated with the release of gaseous halogen
compounds from marine aerosol, mainly in marine and coastal environments.
However, Thornton et al. (2010) found production rates of gaseous nitryl chloride
3-14
-------
near Boulder, Colorado, from reaction of N2O5 with particulate Cl", similar to those
found in coastal and marine environments. C1NO2 readily photolyzes, to yield Cl.
They also found that substantial quantities of N2O5 are recycled through C1NO2 back
into NOX instead of forming HNO3 (a stable reservoir for reactive nitrogen
compounds). The oxidation of hydrocarbons by Cl radicals released from the marine
aerosol could lead to the rapid formation of peroxy radicals and higher rates of O3
production. It should be noted that in addition to production from marine aerosol,
reactive halogen species are also produced by the oxidation of halogenated organic
compounds (e.g., CH3C1, CH3Br, and CH3I). The atmospheric chemistry of halogens
is complex because C1-, Br-, and I-containing species can react among themselves
and with hydrocarbons and other species and could also be important for O3
destruction, as has been noted for the lower stratosphere (McElrov et al.. 1986: Yung
et al.. 1980). For example, the reactions of Br- and Cl-containing radicals deplete O3
in selected environments such as the Arctic during the spring (Barrie et al.. 1988). the
tropical marine boundary layer (Dickerson et al.. 1999). and inland salt flats and salt
lakes (Stutz et al.. 2002). Mahajan et al. (2010) found that I and Br species acting
together resulted in O3 depletion that was much larger than would have been
expected if they acted individually and did not interact with each other; see Annex
AX2.2.10.3 of the 2006 O3 AQCD (U.S. EPA. 2006b).
Multiphase processes have also been associated with the uptake of reactive gas phase
species affecting global budgets of O3 and nitrogen oxides among others. The uptake
of N2O5 on aerosols or cloud droplets leads to the loss of O3 and NOX and the
production of aqueous phase nitric acid, aerosol nitrate, and gaseous halogen nitrites.
In addition to loss of HO2, the uptake of HO2 radicals on aerosol surfaces potentially
reduces O3 concentrations and increases formation of sulfate (if H2O2 is formed after
uptake). Macintyre and Evans (2011) developed a parameterization for uptake of
HO2 based on laboratory studies, which were about a factor of seven lower than
previously estimated. However they note that some of the earlier studies reporting
higher values might have been influenced by transition metal ions (e.g., Cu(II),
Fe(II)), which are highly spatially variable and could be important catalysts in areas
with high concentrations of these ions. Although the global change in O3 was small
(—0.3%) much larger regional changes were found (e.g., up to -27% at the surface
over China).
Uptake coefficients for these species vary widely among laboratory studies.
Macintyre and Evans (2010) showed that the sensitivity of key tropospheric species
such as O3 varies from very small to significant over the range of uptake coefficients
for N2O5 obtained in laboratory studies. For example, global O3 loss ranges from 0
to over 10%, with large regional variability over the range of N2O5 uptake
coefficients reported. In this regard, it should be stressed that knowledge of
multiphase processes is still evolving and there are still many questions that remain
to be answered. However, it is becoming clear that multiphase processes are
important for O3 chemistry.
Reactions of O3 with monoterpenes have been shown to produce oxidants in the
aerosol phase, principally as components of ultrafme particles. Docherty et al. (2005)
3-15
-------
found evidence for the substantial production of organic hydroperoxides in secondary
organic aerosol (SOA) resulting from the reaction of monoterpenes with O3.
Analysis of the SOA formed in their environmental chamber indicated that the SOA
consisted mainly of organic hydroperoxides. In particular, they obtained yields of
47% and 85% of organic peroxides from the oxidation of a- and (3-pinene.
The hydroperoxides then react with aldehydes in particles to form
peroxyhemiacetals, which can either rearrange to form other compounds such as
alcohols and acids or revert back to the hydroperoxides. The aldehydes are also
produced in large measure during the ozonolysis of the monoterpenes. Monoterpenes
also react with OH radicals resulting in the production of more
lower-molecular-weight products than in the reaction with monoterpenes and O3.
Bonn et al. (2004) estimated that hydroperoxides lead to 63% of global SOA
formation from the oxidation of terpenes. The oxidation of anthropogenic aromatic
hydrocarbons by OH radicals could also produce organic hydroperoxides in SOA
(Johnson et al.. 2004). Recent measurements show that the abundance of oxidized
SOA exceeds that of more reduced hydrocarbon like organic aerosol in Pittsburgh
(Zhang et al.. 2005) and in about 30 other cities across the Northern Hemisphere
(Zhang et al.. 2007b). Based on aircraft and ship-based sampling of organic aerosols
over coastal waters downwind of northeastern U.S. cities, de Gouw et al. (2008)
reported that 40-70% of measured organic mass was water soluble and estimated that
approximately 37% of SOA is attributable to aromatic precursors, using PM yields
estimated for NOx-limited conditions. Uncertainties still exist as to the pathways by
which the oxidation of isoprene leads to the formation of SOA. Noziere et al. (2011)
found that a substantial fraction of 2-methyltetrols are primary in origin, although
these species have been widely viewed solely as products of the atmospheric
oxidation of isoprene. This finding points to lingering uncertainty in reaction
pathways in the oxidation of isoprene and in estimates of the yield of SOA from
isoprene oxidation.
Reactions of O3 on the surfaces of particles, in particular those with humic acid like
composition, are instrumental in the processing of SOA and the release of low
molecular weight products such as HCHO (D'Anna et al.. 2009). However, direct
reactions of O3 and atmospheric particles appear to be too slow to represent a major
O3 sink in the troposphere (D'Anna et al.. 2009).
3.2.3.1 Indoor Air
Except when activities such as photocopying or welding are occurring, the major
source of O3 to indoor air is through infiltration of outdoor air. Reactions involving
ambient O3 with NO either from exhaled breath or from gas-fired appliances,
surfaces of furnishings and terpenoid compounds from cleaning products, air
fresheners and wood products also occur in indoor air as was discussed in the 2006
O3 AQCD (U.S. EPA, 2006b). The previous O3 review also noted that the ozonolysis
of terpenoid compounds could be a substantial source of secondary organic aerosol in
the ultrafine size fraction. Chen et al. (2011) examined the formation of secondary
3-16
-------
organic aerosol from the reaction of O3 that has infiltrated indoors with terpenoid
components of commonly used air fresheners. They focused on the formation and
decay of particle bound reactive oxygen species (ROS) and on their chemical
properties. They found that the ROS content of samples can be decomposed into
fractions that differ in terms of reactivity and volatility; however, the overall ROS
content of samples decays and over 90% is lost within a day at room temperature.
This result also suggests loss of ROS during sampling periods longer than a couple of
hours.
3.2.4 Temperature and Chemical Precursor Relationships
As might be expected based on the temperature dependence of many reactions
involved in the production and destruction of O3 and the temperature dependence of
emissions processes such as evaporation of hydrocarbon precursors and the
emissions of biogenically important precursors such as isoprene, ambient
concentrations of O3 also show temperature dependence. Bloomer et al. (2009)
determined the sensitivity of O3 to temperature at rural sites in the eastern United
States. They found that O3 increased on average at rural Clean Air Status and Trends
Network (CASTNET) sites by -3.2 ppbv/°C before 2002; and after 2002 by
~2.2 ppbv/°C. This change in sensitivity was largely the result of reductions in NOX
emissions from power plants. These results are in accord with model predictions by
Wu et al. (2008b) showing that the sensitivity of O3 to temperature decreases with
decreases in precursor emissions. Rasmussen et al. (2012) recently extended the
work of Bloomer et al. (2009) to quantify seasonal changes in the sensitivity of O3 to
temperature as well as regional variability (3-6 ppb/°C over the Northeast and mid-
Atlantic; 3-4 ppb/°C over the Great Lakes region) and to evaluate the capability of a
chemistry-climate model to capture O3 sensitivity to temperature. However, the
associations of O3 with temperature are not as clear in the West as they might be in
the East. For example, sites downwind of Phoenix, AZ showed basically no
sensitivity of O3 to temperature (R2 = 0.02) (U.S. EPA. 2006b). However, Wise and
Comrie (2005) did find that meteorological parameters (mixing height and
temperature) typically accounts for 40 to 70% of the variability in O3 in the five
southwestern cities (including Phoenix) they examined. It is likely that differences in
the nature of sites chosen (urban versus rural) accounted for this difference and are at
least partially responsible for the difference in results. Jaffe et al. (2008) regressed O3
on temperature at Yellowstone and Rocky Mountain NP and found weak associations
(R2 = 0.09 and 0.16). They found that associations with area burned by wildfires are
much stronger. Other sources as discussed in Section 3.4 might also be more
important in the West than in the East.
The warmer months of the year are generally regarded as being the most conducive
to higher O3 concentrations. However, Schnell et al. (2009) reported observations of
high O3 concentrations (maximum 1-h avg of 140 ppb; maximum 8-h avg of
120 ppb) in the Jonah-Pinedale gas fields in Wyoming during winter at temperatures
of-17°C. Potential factors contributing to these anomalously high concentrations
3-17
-------
include a highly reflective snow surface, emissions of NOX, hydrocarbons and short-
lived radical reservoirs (e.g., HONO and HCHO) and a very shallow, stable
boundary layer trapping these emissions close to the surface (Schnell et al., 2009).
Multiphase processes might also be involved in the production of these short-lived
reservoirs. At a temperature of-17°C, the production of hydroxyl radicals (by the
photolysis of O3 yielding OJD followed by the reaction, O(1D) + H2O, needed to
initiate hydrocarbon oxidation) is severely limited, suggesting that another source of
radicals is needed. Radicals can be produced by the photolysis of molecules such as
HONO and HCHO which photolyze in optically thin regions of the solar spectrum.
A similar issue, in part due to the under-prediction of radicals, has arisen in the
Houston airshed where chemistry-transport models (CTMs; discussed further in
Section 3.3) under-predict O3 (Olaguer et al.. 2009). Carter and Seinfeld (2012)
modeled several of the events using the SAPRC-07 chemical mechanism and found
that the release of HONO from the snow surface aids in the formation of O3.
The chemical mechanism they used—including the temperature dependence of rate
coefficients—was developed for application at higher temperatures. They also note
that temperature changes will also affect the distribution of products and radicals
formed when individual VOCs react, but the current version of the mechanism
represents these by lumped overall processes in which the product and radical
distributions are treated as if they are temperature independent. It is not clear how
this treatment of radical production might affect their results.
Rather than varying directly with emissions of its precursors, O3 changes in a
nonlinear fashion with the concentrations of its precursors. At the lowNOx
concentrations found in remote continental areas to rural and suburban areas
downwind of urban centers (low-NOx regime), the net production of O3 typically
increases with increasing NOX. In the low-NOx regime, the overall effect of the
oxidation of VOCs is to generate (or at least not consume) free radicals, and O3
production varies directly with NOX. In the high-NOx regime, NO2 reacts with OH
radicals to form HNO3 (e.g.. Hameed et al.. 1979). These OH radicals would
otherwise oxidize VOCs to produce peroxy radicals, which in turn would oxidize NO
to NO2. In this regime, O3 production is limited by the availability of radicals
(Tonnesen and Jeffries. 1994) and O3 shows only a weak dependence on NOX
concentrations. The production of radicals is in turn limited by the availability of
solar UV radiation capable of photolyzing O3 (in the Hartley bands) or aldehydes
and/or by the abundance of VOCs whose oxidation produce more radicals than they
consume. At the even higher NOX concentrations found in downtown metropolitan
areas, especially near busy streets and roads, and in power plant plumes, there is
scavenging (sometimes referred to as titration) of O3 by reaction with NO to form
NO2 leading to depletion of O3. However, as urban plumes are transported and
diluted, this NO2 can lead to photochemical production of O3 downwind of the
source areas.
The production of radicals can also be limited by the availability of solar UV
radiation capable of photolyzing O3 (in the Hartley bands) and aldehydes. When
solar radiation is blocked by clouds or reduced during winter, VOC and NOX may
3-18
-------
both be available, but O3 production is limited by the availability of solar radiation,
and this has been defined as the "light-limited" regime (Hess et al, 1992).
There are a number of ways to refer to the chemistry of O3 production in these
different chemical regimes. Sometimes the terms VOC-limited and NOx-limited are
used. However, there are difficulties with this usage because (1) VOC measurements
are not as abundant as they are for nitrogen oxides; (2) rate coefficients for reaction
of individual VOCs with radicals (e.g., OH, Cl) vary over an extremely wide range;
and (3) consideration is not given to CH4 or CO and other reactions that can produce
radicals without involving VOCs (e.g., photolysis of HONO). Many of these
difficulties are overcome by the terms NOX -limited and radical-limited (Tonnesen
and Dennis, 2000a, b). This usage recognizes that OH radicals are needed to react
with VOCs to form O3 and that either low NO or high NO2 can limit the production
of OH radicals. This usage also implicitly considers the importance of processes such
as the availability of solar radiation and photolysis in generating radicals. The terms
NOx-limited and NOx-saturated (Jaegle et al., 2001) have also been used to describe
these two regimes. However, the terminology used in original articles will also be
used here. In addition, in the remote marine troposphere, NOX concentrations can be
~20 ppt or less. Under these very low NOX conditions, which are not likely to be
found in the continental U.S., but can characterize inflowing air, HO2 and CH3O2
radicals react with each other and HO2 radicals undergo self-reaction (to form
H2O2), OH radicals efficiently convert NO2 to HNO3, and OH and HO2 react with
O3, leading to net destruction of O3 and inefficient OH radical regeneration.
In addition, halogen-containing radicals also react with O3 acting to keep its
concentrations very low. This is in contrast to the situation in areas of the U.S.
outside of urban cores, where HO2 and CH3O2 radicals react with NO to convert NO
to NO2, regenerate the OH radical, and, through the photolysis of NO2, produce O3
as noted in Annex AX2.2.5 of the 2006 O3 AQCD (U.S. EPA. 2006b).
There are no definitive rules governing the concentrations of NOX at which the
transition from NOX-limited to NOX-saturated conditions occurs. The transition
between these two regimes is highly spatially and temporally dependent and depends
also on the nature and abundance of the hydrocarbons that are present. In a NOX-
limited (or NOx-sensitive) regime, O3 formation is not completely insensitive to
radical production or the flux of solar UV photons, just that O3 formation is more
sensitive to NOX. For example, global tropospheric O3 is sensitive to the
concentration of CH4 even though the troposphere is predominantly NOx-limited.
Likewise, in a NOx-saturated regime there can still be some peroxyl-peroxyl radical
interactions depending on NOX concentrations.
These considerations introduce a high degree of uncertainty into attempts to relate
changes in O3 concentrations to emissions of precursors. The chemistry of OH
radicals, which are responsible for initiating the oxidation of hydrocarbons, shows
behavior similar to that for O3 with respect to NOX concentrations (Poppe et al..
1993: Zimmermann and Poppe. 1993: Hameed et al.. 1979).
Trainer et al. (1993) and Olszynaetal. (1994) have shown that O3 and NOY are
highly correlated in rural areas in the eastern United States. Trainer et al. (1993) also
3-19
-------
showed that O3 concentrations correlate even better with NOZ than with NOY, as
may be expected because NOZ represents the amount of NOX that has been oxidized,
forming O3 in the process. NOZ is equal to the difference between measured total
reactive nitrogen (NOY) and NOX and represents the summed products of the
oxidation of NOX. NOZ is composed mainly of HNO3, PAN and other organic
nitrates, particulate nitrate, and HNO4. Trainer et al. (1993) also suggested that the
slope of the regression line between O3 and NOZ can be used to estimate the rate of
O3 production per NOX oxidized (also known as the O3 production efficiency
[OPE]). Ryerson et al. (2001): Ryerson et al. (1998) used measured correlations
between O3 and NOZ to identify different rates of O3 production in plumes from
large point sources. A number of studies in the planetary boundary layer over the
continental U.S. have found that the OPE ranges typically from 1 to nearly 10.
However, it may be higher in the upper troposphere and in certain areas, such as the
Houston-Galveston area in Texas. Observations indicate that the OPE depends
mainly on the abundance of NOX and also on availability of solar UV radiation,
VOCs and O3 itself.
Various techniques have been proposed to use ambient NOX and VOC
measurements, i.e., as indicators to derive information about the dependence of O3
production on their concentrations. For example, it has been suggested that O3
formation in individual urban areas could be understood in terms of measurements of
ambient NOX and VOC concentrations during the early morning (NRC, 1991).
In this approach, the ratio of summed (unweighted) VOC to NOX is used to
determine whether conditions were NOx-limited or VOC-limited. This approach is
inadequate because it does not consider many factors that are important for O3
production such as the impact of biogenic VOCs (which are typically not present in
urban centers during early morning); important differences in the ability of individual
VOCs to generate radicals (rather than just total VOC) and other differences in O3
forming potential for individual VOCs (Carter. 1995): and changes in the VOC to
NOX ratio due to photochemical reactions and deposition as air moves downwind
from urban areas (Milford et al.. 1994).
Photochemical production of O3 generally occurs simultaneously with the production
of various other species such as HNO3, organic nitrates, and other oxidants such as
hydrogen peroxide. The relative rate of production of O3 and other species varies
depending on photochemical conditions, and can be used to provide information
about O3-precursor sensitivity. Sillman (1995) and Sillman and He (2002) identified
several secondary reaction products that show different correlation patterns for NOX-
limited and NOx-saturated conditions. The most important correlations are for O3
versus NOY, O3 versus NOZ, O3 versus HNO3, and H2O2 versus HNO3.
The correlations between O3 and NOY, and O3 and NOZ are especially important
because measurements of NOY and NOX are more widely available than for VOCs.
Measured O3 versus NOZ (Figure 3-3) shows distinctly different patterns in different
locations. In rural areas and in urban areas such as Nashville, TN, O3 is highly
correlated with NOZ. By contrast, in Los Angeles, CA, O3 is not as highly correlated
with NOZ, and the rate of increase of O3 with NOZ is lower and the O3
concentrations for a given NOZ value are generally lower. The different O3 versus
3-20
-------
NOZ relations in Nashville, TN and Los Angeles, CA reflects the difference between
NOx-limited conditions in Nashville versus an approach to NOx-saturated conditions
in Los Angeles.
10
-x-
X X
20
NOZ (ppb)
30
40
Note: (NOy-NOx) during the afternoon at rural sites in the eastern U.S. (the grey circles) and in urban areas and urban plumes
associated with Nashville, TN (the gray dashes); Paris, France (the black diamonds); and Los Angeles, CA (the Xs).
Source: Adapted with permission of American Geophysical Union, (Sillman and He. 2002: Sillman et al.. 1998: Trainer et al.. 1993).
Figure 3-3 Measured concentrations of O$ and NOz-
The difference between NOx-lirnited and NOx-saturated regimes is also reflected in
measurements of H2O2. H2O2 production is highly sensitive to the abundance of
radicals and is thus favored in the NOx-limited regime. Measurements in the rural
eastern U.S. (Jacob etal.. 1995). Nashville, TN (Sillman et al.. 1998). and
Los Angeles, CA (Sakugawa and Kaplan. 1989). show large differences in H2O2
concentrations between likely NOx-limited and NOx-saturated locations. Tonnesen
and Dennis (2000a) reviewed the use of many indicator ratios and Tonnesen and
Dennis (2000b) proposed HCHO/NO2 as an indicator to distinguish between NOX-
limited and radical-limited regimes.
The applications of indicator species mentioned above are mainly limited to
individual urban areas because these are the areas where it is often not clear which
regime is prevalent. Satellites provide a platform for greatly extending the range of
applicability of the indicator technique and also have the resolution necessary to
examine urban to rural differences. Martin et al. (2004) and Duncan et al. (2010)
used satellite data from Ozone Monitoring Instrument (OMI) for HCHO to NO2
column ratios to diagnose NOx-limited and radical-limited (NOx-saturated) regimes.
HCHO can be used as an indicator of VOCs as it is a common, short-lived, oxidation
3-21
-------
product of many VOCs that is a source of HOX (Sillman, 1995). In adopting the
satellite approach, CTMs are used to estimate the fractional abundance of the
indicator species in the planetary boundary layer. Duncan et al. (2010) found that O3
formation over most of the U.S. became more sensitive to NOX over most of the U.S.
from 2005 to 2007 largely because of decreases in NOX emissions. They also found
that surface temperature is correlated with the ratio of HCHO to NO2 especially in
cities in the Southeast where emissions of isoprene (a major source of HCHO) are
high due to high temperatures in summer.
3.3 Atmospheric Modeling
CTMs have been widely used to compute the interactions among atmospheric
pollutants and their transformation products, and the transport and deposition of
pollutants. They have also been widely used to improve basic understanding of
atmospheric chemical processes and to develop control strategies. The spatial scales
over which pollutant fields are calculated range from intra-urban to regional to
global. Generally, these models are applied to problems on different spatial scales but
efforts are underway to link across spatial scales for dealing with global scale
environmental issues that affect population health within cities. Many features are
common to all of these models and hence they share many of the same problems.
On the other hand, there are appreciable differences in approaches to parameterizing
physical and chemical processes that must be addressed in applying these models
across spatial scales.
CTMs solve a set of coupled, non-linear partial differential equations, or continuity
equations, for relevant chemical species. Jacobson (2005) described the governing
partial differential equations, and the methods that are used to solve them. Because of
limitations imposed by the complexity and spatial-temporal scales of relevant
physical and chemical processes, the CTMs must include parameterizations of these
processes, which include atmospheric transport; the transfer of solar radiation
through the atmosphere; chemical reactions; and removal to the surface by turbulent
motions and precipitation. Development of parameterizations for use in CTMs
requires data for three dimensional wind fields, temperatures, humidity, cloudiness,
and solar radiation; emissions data for primary (i.e., directly emitted from sources)
species such as NOX, SO2, NH3, VOCs, and primary PM; and chemical reactions.
The domains of CTMs extend from a few hundred kilometers on a side to the entire
globe. Most major regional (i.e., sub-continental) scale air-related modeling efforts at
EPA rely on the Community Multi-scale Air Quality (CMAQ) modeling system
(Byun and Schere, 2006; Byun and Ching, 1999). CMAQ's horizontal domain
typically extends over North America with efforts underway to extend it over the
entire Northern Hemisphere. Note that CTMs can be 'nested' within each other as
shown in Figure 3-4 which shows domains for CMAQ (Version 4.6.1); additional
details on the model configuration and application are found elsewhere (U.S. EPA,
2009e). The figure shows the outer domain (36 km horizontal grid spacing) and two
3-22
-------
12 km spatial resolution (east and west) sub-domains. The upper boundary for
CMAQ is typically set at about 100 hPa, or at about 16 km altitude on average,
although in some recent applications the upper boundary has been set at 50 hPa.
These domains and grid spacings are quite common and can also be found in a
number of other models.
Note: This figure depicts a 36 km grid-spacing outer parent domain in black; 12 km western U.S. domain in red; 12 km
eastern U.S. domain in blue.
Figure 3-4 Sample Community Multi-scale Air Quality (CMAQ) modeling
domains.
The main components of a CTM such as EPA's CMAQ are summarized in
Figure 3-5. The capabilities of a number of CTMs designed to study local- and
regional-scale air pollution problems were summarized by Russell and Dennis (2000)
and in the 2006 O3 AQCD (U.S. EPA. 2006b). Historically, CMAQ has been driven
most often by the MM5 mesoscale meteorological model (Seaman, 2000), though it
could be driven by other meteorological models including the Weather Research and
Forecasting (WRF) model and the Regional Atmospheric Modeling System (RAMS)
(ATMET. 2011).
3-23
-------
Chemistry Transport Model
Visualization of Output
Process Analyses
Figure 3-5 Main components of a comprehensive atmospheric chemistry
modeling system, such as the U.S. EPA's Community Multi-scale
Air Quality (CMAQ) modeling system.
Simulations of pollution episodes over regional domains have been performed with a
horizontal resolution down to 1 km; see the application and general survey results
reported in Ching et al. (2006). However, simulations at such high resolution require
better parameterizations of meteorological processes such as boundary layer fluxes,
deep convection, and clouds (Seaman, 2000). Finer spatial resolution is necessary to
resolve features such as urban heat island circulation; sea, bay, and land breezes;
mountain and valley breezes; and the nocturnal low-level j et; all of which can affect
pollutant concentrations. Other major air quality systems used for regional scale
applications include the Comprehensive Air Quality Model with extensions (CAMx)
(ENVIRON, 2005) and the Weather Research and Forecast model with Chemistry
(WRF/Chem) (NOAA. 2010).
CMAQ and other grid-based or Eulerian air quality models subdivide the modeling
domain into a three-dimensional array of grid cells. The most common approach to
setting up the horizontal domain is to nest a finer grid within a larger domain of
coarser resolution. The use of finer horizontal resolution in CTMs will necessitate
finer-scale inventories of land use and better knowledge of the exact paths of roads,
locations of factories, and, in general, better methods for locating sources and
estimating their emissions. The vertical resolution of CTMs is variable and usually
configured to have more layers in the PEL and fewer in the free troposphere.
3-24
-------
The meteorological fields are produced either by other numerical prediction models
such as those used for weather forecasting (e.g., MM5, WRF), and/or by assimilation
of satellite data. The flow of information shown in Figure 3-5 has most often been
unidirectional in the sense that information flows into the CTM (large box) from
outside; feedbacks on the meteorological fields and on boundary conditions (i.e., out
of the box) have not been included. However, CTMs now have the capability to
consider these feedbacks as well; see, for example, Binkowski et al. (2007) and
WRF/Chem (NOAA. 2010).
Because of the large number of chemical species and reactions that are involved in
the oxidation of realistic mixtures of anthropogenic and biogenic hydrocarbons,
condensed mechanisms must be used in atmospheric models. These mechanisms can
be tested by comparison with smog chamber data. However, the existing chemical
mechanisms often neglect many important processes such as the formation and
subsequent reactions of long-lived carbonyl compounds, the incorporation of the
most recent information about intermediate compounds, and heterogeneous reactions
involving cloud droplets and aerosol particles. To the extent that information is
available, models like CMAQ and CAMx do include state-of-the-science
parameterization for some of these processes such as heterogeneous N2O5 chemistry.
The initial conditions, or starting concentration fields of all species computed by a
model, and the boundary conditions, or concentrations of species along the horizontal
and upper boundaries of the model domain throughout the simulation, must be
specified at the beginning of the simulation. Both initial and boundary conditions can
be estimated from models or data or, more generally, model plus data hybrids.
Because data for vertical profiles of most species of interest are very sparse, results
of model simulations over larger, usually global, domains are often used.
Chemical kinetics mechanisms representing the important reactions occurring in the
atmosphere are used in CTMs to estimate the rates of chemical formation and
destruction of each pollutant simulated as a function of time. The Master Chemical
Mechanism (MCM) (Univ of Leeds. 2010) is a comprehensive reaction database
providing as near an explicit treatment of chemical reactions in the troposphere as is
possible. The MCM currently includes over 12,600 reactions and 4,500 species.
However, mechanisms that are this comprehensive are still computationally too
demanding to be incorporated into CTMs for regulatory use. Simpler treatments of
tropospheric chemistry have been assembled by combining chemical species into
mechanisms that group together compounds with similar chemistry. It should be
noted that because of different approaches to the lumping of organic compounds into
surrogate groups for computational efficiency, chemical mechanisms can produce
different results under similar conditions. Jimenez et al. (2003) briefly described the
features of the seven main chemical mechanisms in use and compared concentrations
of several key species predicted by these mechanisms in a box-model simulation over
24 hours. Several of these mechanisms have been incorporated into CMAQ including
extensions of the Carbon Bond (CB) mechanism (Luecken et al.. 2008). SAPRC
(Luecken et al.. 2008). and the Regional Atmospheric Chemistry Mechanism, version
2 (RACM2) (Fuentes et al.. 2007). The CB mechanism is currently undergoing
3-25
-------
extension (CB06) to include, among other things, longer lived species to better
simulate chemistry in the remote and upper troposphere. These mechanisms were
developed primarily for homogeneous gas phase reactions and treat multiphase
chemical reactions in a very cursory manner, if at all. As a consequence of neglecting
multiphase chemical reactions, models such as CMAQ could have difficulties
capturing the regional nature of O3 episodes, in part because of uncertainty in the
chemical pathways converting NOX to HNO3 and recycling of NOX (Godowitch et
al.. 2008: Hains et al.. 2008). Much of this uncertainty also involves multiphase
processes as described in Section 3.2.3.
CMAQ and other CTMs incorporate processes and interactions of aerosol-phase
chemistry (Zhang and Wexler, 2008; Gaydos et al., 2007; Binkowski and Roselle,
2003). There have also been several attempts to study the feedbacks of chemistry on
atmospheric dynamics using meteorological models like MM5 and WRF (Liu et al.,
2001; Park et al.. 2001; Grell et al.. 2000; Luetal.. 1997). This coupling is necessary
to accurately simulate feedbacks from PM (Park et al., 2001; Lu et al., 1997) over
areas such as Los Angeles or the Mid-Atlantic region. Photolysis rates in CMAQ can
now be calculated interactively with model produced O3, NO2, and aerosol fields
(Binkowski et al.. 2007).
Spatial and temporal characterizations of anthropogenic and biogenic precursor
emissions can be specified as inputs to a CTM or these emissions can be calculated
in-line in CMAQ. Emissions inventories have been compiled on grids of varying
resolution for many hydrocarbons, aldehydes, ketones, CO, NH3, and NOX.
Preprocessing of emissions data for CMAQ is done by the Spare-Matrix Operator
Kernel Emissions (SMOKE) system (UNC. 2011). For many species, information on
temporal variability of emissions is lacking, so long-term annual averages are used in
short-term, episodic simulations. Annual emissions estimates can be modified by the
model to produce emissions more characteristic of the time of day and season.
Appreciable errors in emissions can occur if inappropriate time dependence is
applied.
Each of the model components described above has associated uncertainties; and the
relative importance of these uncertainties varies with the modeling application. Large
errors in photochemical modeling arise from the meteorological, chemical and
emissions inputs to the model (Russell and Dennis. 2000). While the effects of
poorly specified boundary conditions propagate through the model's domain, the
effects of these errors remain undetermined. Because many meteorological processes
occur on spatial scales smaller than the model's vertical or horizontal grid spacing
and thus are not calculated explicitly, parameterizations of these processes must be
used. These parameterizations introduce additional uncertainty.
The performance of CTMs must be evaluated by comparison with field data as part
of a cycle of model evaluations and subsequent improvements (NRC, 2007).
However, they are too computationally demanding to have the full range of their
sensitivities examined using Monte Carlo techniques (NRC, 2007). Models of this
complexity are evaluated by comparison with field observations for O3 and other
species. Evaluations of the performance of CMAQ are given in Arnold et al. (2003),
3-26
-------
Eder and Yu (2005). Appel et al. (2005). and Fuentes and Raftery (2005).
Discrepancies between model predictions and observations can be used to point out
gaps in current understanding of atmospheric chemistry and to spur improvements in
parameterizations of atmospheric chemical and physical processes. Model evaluation
does not merely involve a straightforward comparison between model predictions
and the concentration field of the pollutant of interest. Such comparisons may not be
meaningful because it is difficult to determine if agreement between model
predictions and observations truly represents an accurate treatment of physical and
chemical processes in the CTM or the effects of compensating errors in complex
model routines (in other words, it is important to know if the right answer is obtained
for the right reasons). Each of the model components (emissions inventories,
chemical mechanism, and meteorological driver) should be evaluated individually as
has been done to large extent in some major field studies such as TexAQS I and II
and CalNex. Comparisons of correlations between measured and modeled VOCs and
NOX are useful for evaluating results from CTMs and can provide information about
the chemical state of the atmosphere. A CTM that accurately computes both VOC
and NOX along with the spatial and temporal relations among the critical secondary
species associated with O3 has a higher probability of representing O3-precursor
relations correctly than one that does not.
The above evaluation techniques are sometimes referred to as "static" in the sense
that individual model variables are compared to observations. It is also crucial to
understand the dynamic response to changes in inputs and to compare the model
responses to those that are observed. These tests might involve changes in some
natural forcing or in emissions from an anthropogenic source. As an example,
techniques such as the direct decoupled method (DDM) could be used in CTMs to
determine their first order sensitivity to emissions changes (Zhang. 2005: Dunker et
al.. 2002). However, the observational basis for comparing a model's response is
largely unavailable for many problems of interest, in large part because
meteorological conditions are also changing while the emissions are changing. As a
result, methods such as DDM are used mainly to assess the potential effectiveness of
emissions controls (Arunachalam. 2009). Because the chemistry of O3 formation is
non-linear near strong sources, and higher order terms taking into account the
curvature of the response surface of O3 with respect to changes in sources are needed
when using DDM. These additional terms are incorporated in the higher order
decoupled method, or HDDM, (see the U.S. EPA technical memorandum on
applications of HDDM, particularly with respect to adjusting O3 concentrations in
response to emissions reductions [or "roll back"] at
http://www.epa. gov/ttnnaaqs/standards/ozone/s_o3_2008_td.html).
3.3.1 Global Scale CTMs
With recognition of the global nature of many air pollution problems, global scale
CTMs have been applied to regional scale pollution problems (NRC. 2009). Global-
scale CTMs are used to address issues associated with global change, to characterize
3-27
-------
long-range transport of air pollutants, and to provide boundary conditions for the
regional-scale models. The upper boundaries of global scale CTMs extend anywhere
from the tropopause (~8 km at the poles to -16 km in the tropics) to the mesopause at
~80 km, in order to obtain more realistic boundary conditions for problems involving
stratospheric dynamics and chemistry. The global-scale CTMs consider the same
processes shown in Figure 3-5 for the regional scale models. In addition, many of the
same issues that have arisen for the regional models have also arisen for the global
scale models (Emmerson and Evans. 2009). For example, after adjusting lightning
NOX to better match observed constraints in the MOZART-4 model, simulated
HNO3 was too low and PAN too high in the mid-troposphere, though observations
were captured in the upper troposphere, over the U.S. during summer 2004 in the
MOZART-4 model (Fang et al.. 2010). In contrast, summer 2004 simulations with
improved lightning NOX in GEOS-Chem indicate that PAN is too low but HNO3 is
overestimated throughout the mid- and upper troposphere (Hudman et al.. 2007).
Predictions of HNO3 were too high and PAN too low over the U.S. during summer in
the MOZART model (Fang et al.. 2010). Similar findings were obtained in a box
model of upper tropospheric chemistry (Henderson et al.. 2011). indicating a need for
improved constraints on processes controlling NOY distributions in the free
troposphere.
The GEOS-Chem model is a community-owned, global scale CTM that has been
widely used to study issues associated with the hemispheric transport of pollution
and global change (Harvard University, 2010a). Comparisons of the capabilities of
GEOS-Chem and several other models to simulate intra-hemispheric transport of
pollutants are given in a number of articles (Fiore et al., 2009; Reidmiller et al.,
2009). Reidmiller et al. (2009) compared the ensemble average of 18 global models
to spatially and monthly averaged observations of O3 at CASTNET sites in the U.S.
(see Figure 3-6). These results show that the multi-model ensemble agrees much
better with observations than do most of the individual models. The GEOS-Chem
model was run for two grid spacings (4° x 4.5° and 2° x 2.5°) over the U.S. with very
similar results that lie close to the ensemble average. In general, the model ensemble
average and the two GEOS-Chem simulations are much closer to observations in the
Intermountain West Region than in the Southeast Region during summer, when most
major O3 episodes occur (Note, though, that more current versions of GEOS-Chem
are now in use.) However, there are also sizable over-predictions by many models in
both regions during summer.
3-28
-------
Mountain West Region
i—i—i—i—i—i—i—i—i
CAMCHEM
ECHAM5
EWEP
FRSGCUCI
GEMAQ-vl pO
GEOSChem-v07
GEOSChem-w4S
GISS-PUCCINI
GWI
MCA-vSSz
LLNL-IWPACT
MOZARTGFDL
MOZECH
OsloCTMZ
TM5-JRC
DBS
Multi-mod si mean
F M A M J
A S O N D
Source: Reprinted with permission of Copernicus Publications, (Reidmiller et al.. 2009).
Figure 3-6 Comparison of global chemical-transport model (CTM) predictions
of daily maximum 8-h avg Os concentrations and multi-model mean
with monthly averaged CASTNET observations in the
Intermountain West and Southeast Regions of the U.S.
In their review, McDonald-Buller et al. (2011) noted that global scale chemical
transport models exhibit biases in monthly mean of the daily maximum 8-h avg
(MDA8) O3 in some regions of the U.S., including the Gulf Coast, regions affected
by fires, and regions with complex topography, which have implications for model
estimates of background O3; and they also have difficulty representing the fine
structures of O3 events at sub-grid scales at relatively remote monitoring sites that
include contributions to O3 from background sources.
Global models are not alone in overestimating O3 in the Southeast. Godowitch et al.
(2008). Gilliland et al. (2008) and Nolte et al. (2008) found positive O3 biases in
regional models over the eastern U.S., as well, which they largely attributed to
uncertainties in temperature, relative humidity and planetary boundary layer height.
Agreement between monthly average values is expected to be better than with daily
values because of a number of factors including the increasing uncertainty of
emissions at finer time resolution. Kasibhatla and Chameides (2000) found that the
3-29
-------
accuracy of simulations improved as the averaging time of both the simulation and
the observations increased.
Simulations of the effects of long-range transport at particular locations must be able
to link multiple horizontal resolutions from the global to the local scale. Because of
computational limitations, global simulations are not made at the same horizontal
resolutions found in the regional scale models, i.e., down to 1-4 km2 horizontal
resolution. They are typically conducted with a horizontal grid spacing of l°-2° of
latitude and longitude (or roughly 100-200 km at mid-latitudes). Some models such
as GEOS-Chem have the capability to include nested models at a resolution of
0.5° x 0.667° (Wang et al., 2009a) and efforts are underway to achieve even higher
spatial resolution. Another approach is to nest regional models within GEOS-Chem.
Caution must be exercised with nesting different models because of differences in
chemical mechanisms and numerical schemes, and in boundary conditions between
the outer and inner models. As an example of these issues, surface O3 concentrations
that are too high have been observed in models in which CMAQ was nested inside of
GEOS-Chem. The high O3 results in large measure from stratospheric O3 intruding
into the CMAQ domain [see (Lam and Fu, 2010) for one way to address this issue].
Large vertical O3 gradients in the upper troposphere must be preserved to accurately
represent downward transport of stratospheric O3. This complicates efforts to link
global and regional models with different vertical grid spacing. Efforts are also
underway to extend the domain of CMAQ over the entire Northern Hemisphere.
In this approach, the same numerical schemes are used for transporting species and
the same chemistry is used throughout all spatial scales. Finer resolution in models of
any scale can only improve scientific understanding to the extent that the governing
processes are accurately described. Consequently, there is a crucial need for
observations at the appropriate scales to evaluate the scientific understanding
represented by the models.
3.4 Background O3 Concentrations
Background concentrations of O3 have been given various definitions in the literature
over time. An understanding of the sources and contributions of background O3 to
O3 concentrations in the U.S. is potentially useful in reviewing the O3 NAAQS,
especially related to days at the upper end of the distribution of O3 concentrations.
In the context of a review of the NAAQS, it is useful to define background O3
concentrations in a way that distinguishes between concentrations that result from
precursor emissions that are relatively less controllable from those that are relatively
more controllable through U.S. policies. In previous NAAQS reviews, a specific
definition of background concentrations was used and referred to as policy relevant
background (PRB). In those previous reviews, PRB concentrations were defined by
EPA as those concentrations that would occur in the U.S. in the absence of
anthropogenic emissions in continental North America (CNA), defined here as the
U.S., Canada, and Mexico. There is no chemical difference between background O3
and O3 attributable to CNA anthropogenic sources. However, to inform policy
3-30
-------
considerations regarding the current or potential alternative standards, it is useful to
understand how total O3 concentrations can be attributed to different sources.
For this document, EPA has considered background O3 concentrations more broadly
by considering three different definitions of background. The first is natural
background which includes contributions resulting from emissions from natural
sources (e.g., stratospheric intrusion, wildfires, biogenic methane, and more short-
lived VOC emissions) throughout the globe simulated in the absence of all
anthropogenic emissions. The second is North American background
(NA background) which includes contributions from natural background throughout
the globe and emissions of anthropogenic pollutants contributing to global
concentrations of O3 (e.g., anthropogenic methane) from countries outside North
America. The third is United States background (U.S. background) which includes
contributions from natural background throughout the globe and emissions from
anthropogenic pollutants contributing to global concentrations of O3 from countries
outside the United.States. U.S. background differs from NA background in that it
includes anthropogenic emissions from neighboring Canada and Mexico. These three
definitions have been explored in recent literature and are discussed further below.
Sources included in the definitions of NA-background and U.S.-background O3 are
shown schematically in Figure 3-7. Definitions of background and approaches to
derive background concentrations were reviewed in the 2006 O3 AQCD (U.S. EPA.
2006b) and in Reid et al. (2008). Further detail about the processes involved in these
sources is given in Section 3.4.1 and Section 3.4.2 and application to models
calculating background concentrations is presented in Section 3.4.3.
3-31
-------
Stratosphere
Outside natural
influences
Lightning
Long-range transport
of pollution
"Background" air
Fires Land Human
biosphere activity
Note: Background concentrations are O3 concentrations that would exist in the absence of anthropogenic emissions from the U.S.,
Canada, and Mexico. United States (U.S.) background is similarly defined, but includes transport from Canada and Mexico in
addition to intercontinental transport.
Figure 3-7 Schematic overview of contributions to North American
(NA) background concentrations of O$.
3.4.1 Contributions from Natural Sources
Natural sources contributing to background O3 include the stratospheric-tropospheric
exchange (STE) of O3 and photochemical reactions involving natural O3 precursor
emissions of VOCs, NOX, and CO. Natural sources of O3 precursors include
biogenic emissions, wildfires, and lightning. Biogenic emissions from agricultural
activities in CNA (or the U.S.) are not considered in the formation of NA- (or
U.S.)-background O3. Contributions from natural sources are an important
component of background concentrations and are discussed in greater detail below.
3.4.1.1 Contributions from the Stratosphere
The basic atmospheric dynamics and thermodynamics of STE were outlined in the
2006 O3 AQCD (U.S. EPA. 2006b): as noted there, stratospheric air rich in O3 is
transported into the troposphere. Ozone is produced naturally by photochemical
reactions in the stratosphere as shown in Figure 3-1. Some of this O3 is transported
downward into the troposphere throughout the year, with maximum contributions at
mid-latitudes during late winter and early spring mainly coming from a process
known as tropopause folding. These folds occur behind most cold fronts, bringing
3-32
-------
stratospheric air with them. The tropopause should not be interpreted as a material
surface through which there is no exchange. Rather these folds should be thought of
as regions in which mixing of tropospheric and stratospheric air is occurring
(Shapiro, 1980). This imported stratospheric air contributes to the natural background
of O3 in the troposphere, especially in the free troposphere during winter and spring.
Significant intrusions of stratospheric air occur in "ribbons" -200 to 1000 km in
length, 100 to 300 km wide and about 1 to 4 km thick (Wimmers et al.. 2003:
Hoskins. 1972). Thus, these intrusions are large scale three-dimensional events and
should not be thought of as one-dimensional. STE also occurs during other seasons
including summer.
Methods for estimating the contribution of stratospheric intrusions rely on the use of
tracers of stratospheric origin that can be either dynamical or chemical. Thompson et
al. (2007) found that roughly 20-25% of free tropospheric O3 over northeastern
North America during July-August 2004 was of stratospheric origin based on an
analysis of ozonesonde data. This O3 can be mixed into the PEL where it can either
be destroyed or transported to the surface. They relied on the combined use of low
relative humidity and high (isentropic) potential vorticity (PV) (>2 PV units) to
identify stratospheric contributions. PV has been a widely used tracer for
stratospheric air; see the 2006 O3 AQCD (U.S. EPA, 2006b). However, bear in mind
that this analysis pertains to a single year only with no indication of interannual
variability. Lefohn et al. (2011) used these and additional criteria to assess
stratospheric influence on sites in the Intermountain West and in the Northern Tier.
Additional criteria include consideration of trajectories originating at altitudes above
the 380 K potential temperature surface with a residence time requirement at these
heights. Based on these criteria, they identified likely stratospheric influence at the
surface sites on a number of days during spring of 2006 to 2008. However, they
noted that their analysis of stratospheric intrusions captures only the frequency and
vertical penetration of the intrusions but does not provide information about the
contribution of the intrusions to the measured O3 concentration. These results are all
generally consistent with what was noted in the 2006 O3 AQCD (U.S. EPA. 2006b).
Fischer (2004) analyzed the O3 record during summer at Mount Washington and
identified a stratospheric contribution to 5% of events during the summers of 1998 -
2003 when O3 was >65 ppb; the air was dry and trajectories originated from altitudes
where PV was elevated (PV >1 PV unit). However, this analysis did not quantify the
relative contributions of anthropogenic and stratospheric O3 sources more precisely,
because as they note identifying stratospheric influences is complicated by transport
over industrialized/urban source regions. Stratospheric O3 was hypothesized to
influence the summit during conditions also potentially conducive to photochemical
O3 production, which make any relative contribution calculations difficult without
additional measurements of anthropogenic and stratospheric tracers.
Although most research has been conducted on tropopause folding as a source of
stratosphere to troposphere exchange, this is not the only mechanisms by which
stratospheric O3 can be brought to lower altitudes. Tang et al. (2011) estimated that
deep convection capable of penetrating the tropopause can increase the overall
downward flux of O3 by -20%. This mechanism operates mainly during summer in
3-33
-------
contrast with tropopause folding which is at a maximum from late winter through
spring and at lower latitudes. Yang et al. (2010) estimated that roughly 20% of free
tropospheric O3 above coastal California in 2005 and 2006 was stratospheric in
origin. Some of this O3 could also contribute to O3 at the surface.
It should be noted that there is considerable uncertainty in the magnitude and
distribution of this potentially important source of tropospheric O3. Stratospheric
intrusions that reach the surface are much less frequent than intrusions which
penetrate only to the middle and upper troposphere. However, O3 transported to the
upper and middle troposphere can still affect surface concentrations through various
exchange mechanisms that mix air from the free troposphere with air in the PEL.
Several instances of STE producing high concentrations of O3 around Denver and
Boulder, Colorado, were analyzed by Langford et al. (2009) and several likely
instances of STE, including one of the cases analyzed by Langford et al. (2009) were
also cited in Annex AX2-3 of the 2006 O3 AQCD (U.S. EPA. 2006b). Clear
examples of STE have also been observed in southern Quebec province by Hocking
et al. (2007). in accord with previous estimates by Wernli and Bourqui (2002) and
James et al. (2003). As also noted in the 2006 O3 AQCD (U.S. EPA. 2006b). the
identification of stratospheric O3 and the calculation of its contributions to ambient
air requires data for other tracers of stratospheric origin. In some cases, stratospheric
O3 intrusions can be identified based on measurements of low relative humidity, high
potential vorticity and low ratios of O3/PM. Strong stratospheric O3 intrusion events
that typically occur during winter or spring have been readily identified using these
types of data (Langford et al.. 2009). However, it remains challenging to accurately
estimate the contributions from smaller direct or indirect (i.e., resulting from shallow
intrusions into the mid and upper troposphere that are then mixed downward into the
planetary boundary layer) contributions of stratospheric O3 to ambient air.
3.4.1.2 Contributions from Other Natural Sources
Biomass burning consists of wildfires and the intentional burning of vegetation to
clear new land for agriculture and for population resettlement; to control the growth
of unwanted plants on pasture land; to manage forest resources with prescribed
burning; to dispose of agricultural and domestic waste; and as fuel for cooking,
heating, and water sterilization. Biomass burning also exhibits strong seasonality and
interannual variability (van der Werf et al.. 2006). with most biomass burned during
the local dry season. This is true for both prescribed burns and wildfires. Globally,
most wildfires may be ignited directly as the result of human activities, leaving only
10-30% initiated by lightning (Andreae. 1991). However, because fire management
practices suppress natural wildfires, the buildup of fire fuels increases the
susceptibility of forests to more severe but less frequent fires. Thus there is
considerable uncertainty in attributing the fraction of wildfire emissions to human
activities because the emissions from naturally occurring fires that would have been
present in the absence of fire suppression practices are not known. Contributions to
3-34
-------
NOX, CO, and VOCs from wildfires and prescribed fires are considered as precursors
to background O3 formation in this assessment.
Estimating contributions from wildfires is subject to considerable uncertainty.
McDonald-Buller et al. (2011) note that "Models generally find little O3 production
in wildfire plumes for short aging times (days) because NOX emissions are low and
conversion to peroxyacetylnitrate (PAN) is rapid. In contrast, observations show
large O3 production from at least some regional wildfires that may significantly
elevate O3 at low altitude sites on a monthly basis, and persist over long distances
from the burned region." They also note that fire plumes transported on
intercontinental scales can contain very high O3 concentrations. However Singh et al.
(201 Ob) found appreciable increases of O3 in California fire plumes only when they
are mixed with urban pollution. Jaffe and Wigder (2012) note that this result could
have also been due to suppression of O3 production near the source. Indeed, Singh et
al. (2012) note that low O3 found in the California fire plumes of 2008 were
associated with low concentrations of NOX and that NOX from urban emissions was
needed to produce high O3. Factors such as the stage of combustion (smoldering to
flaming), fuel nitrogen content, ambient meteorological conditions, and the
availability of solar ultraviolet radiation need to be considered when evaluating the
potential of fires for producing O3.
Jaffe et al. (2008) examined the effects of wildfires on O3 in the western United
States. They found a strong relation (R2 = 0.60) between summer mean O3 measured
at various national park and CASTNET sites and area burned in the western United
States. They also found generally higher concentrations within surrounding
5° x 5° and 10° x 10° of burned areas. Smaller correlations were found within the
surrounding l°x 1° areas, reflecting near source consumption of O3 and the time
necessary for photochemical processing of emissions to form O3. Jaffe et al. (2008)
estimate that burning 1 million acres in the western U.S. during summer results in an
increase in O3 of 2 ppb across the region; this translates to an average O3 increase
across the entire western U.S. of 3.5 and 8.8 ppb during mean and maximum fire
years. The unusually warm and dry weather in central Alaska and western Yukon in
the summer of 2004 contributed to the burning of 11 million acres there. Subsequent
modeling by Pfister et al. (2005) showed that the CO contribution from these fires in
July 2004 was 33.1 (±5.5) MT that summer, roughly comparable to total U.S.
anthropogenic CO emissions during the same period.
These results underscore the importance of wildfires as a source of important O3
precursors. In addition to emissions from forest fires in the U.S., emissions from
forest fires in other countries can be transported to the U.S., for example from boreal
forest fires in Canada (Mathur. 2008). Siberia (Generoso et al.. 2007) and tropical
forest fires in the Yucatan Peninsula and Central America (Wang et al.. 2006). These
fires have all resulted in notable increases in O3 concentrations in the U.S.
Estimates of biogenic VOC, NO, and CO emissions can be made using the BEIS
model with data from the BELD and annual meteorological data or MEGAN. VOC
emissions from vegetation were described in Section 3.2.
3-35
-------
As discussed in Section 3.2.1, NOX is produced by lightning. Kaynak et al. (2008)
found lightning contributes 2 to 3 ppb to surface-level background O3 centered
mainly over the southeastern U.S. during summer. Although total column estimates
of lightning produced NOX are large compared to anthropogenic NOX during
summer, lightning produced NOX does not contribute substantially to the NOX
burden in the continental boundary layer. For example, (Fang et al.. 2010) estimated
that only 2% of NOX production by lightning occurs within the boundary layer and
most occurs in the free troposphere. In addition, much of the NOX produced in the
free troposphere is converted to more oxidized N species during downward transport.
Note that contributions of natural sources to North American background arise from
everywhere in the world.
3.4.2 Contributions from Anthropogenic Emissions
In addition to emissions from North America, anthropogenic emissions from Eurasia
have contributed to the global burden of O3 in the atmosphere and to the U.S. (NRC,
2009, and references therein). Because the mean tropospheric lifetime of O3 is on the
order of a few weeks (Hsu and Prather, 2009), O3 can be transported from continent
to continent and around the globe in the Northern Hemisphere. Ozone produced by
U.S. emissions can, therefore, be recirculated around northern mid-latitudes back to
the United States. High elevation sites are most susceptible to the intercontinental
transport of pollution especially during spring. For example, a number of occurrences
of O3>60 ppb from mid-April to mid-May of 2006 were observed at Mt. Bachelor
Observatory, OR (elevation 2,700 meters) with a maximum O3 concentration of
~85 ppb observed on April 22, 2006. Calculations using GEOS-Chem, a global-scale
CTM, indicate that Asia contributed 9 ± 3 ppb to a modeled mean concentration of
53 ± 9 ppb O3 at Mt. Bachelor during the same period compared to measured
concentrations of 54 ± 10 ppb (Zhang et al., 2008). Zhang et al. (2008) also
calculated a contribution of 5 to 7 ppb to surface O3 over the western U.S. during
that period from Asian anthropogenic emissions. They also estimated an increase in
NOX emissions of-44% from Asia from 2001 to 2006 resulting in an increase of 1-
2 ppb in O3 over North America.
Cooper et al. (2010) analyzed all available O3 measurements in the free troposphere
above western North America at altitudes of 3-8 km (above sea level) during April
and May of 1995 to 2008 (i.e., times when intercontinental transport is most
prominent). They derived a trend of+0.63 ± 0.34 ppb/year in median O3
concentrations with indication of a similar rate of increase since 1984. Back
trajectories that were likely to have been strongly and recently influenced by North
American emissions were filtered out, resulting in a trend of+0.71 ± 0.45 ppb/year.
Considering only trajectories with an Asian origin resulted in a trend of
+0.80 ± 0.34 ppb/year. These results suggest that local North American emissions
were not responsible for the measured O3 increases. This O3 could have been
produced from natural and anthropogenic precursors in Asia and Europe with some
contribution from North American emissions that have circled the globe. Cooper et
3-36
-------
al. (2010) also found that it is unlikely that the trends in tropospheric O3 are
associated with trends in stratospheric intrusions. Note, however, that these results
relate to O3 trends above ground level and not to surface O3. Model results (Zhang et
al., 2008) show that surface O3 contributions from Asia are much smaller than those
derived in the free troposphere because of dilution and chemical destruction during
downward transport to the surface. These processes tend to reduce the strength of
associations between free tropospheric and surface O3 especially if air from other
sources is sampled by the surface monitoring sites.
Trinidad Head, CA is one sampling location at which measurements might be
expected to reflect in large measure NA background O3 contributions, at times
during the spring (Oltmans et al., 2008; Goldstein et al., 2004). The monitoring
station at Trinidad Head is on an elevated peninsula extending out from the mainland
of northern California, and so might be expected at times to intercept air flowing in
from the Pacific Ocean with little or no influence from sources on the mainland.
Figure 3-8 shows the time series of MDA8 O3 concentrations measured at Trinidad
Head from April 18, 2002 through December 31, 2009. The data show pronounced
seasonal variability with spring maxima and summer minima. Springtime
concentrations typically range from 40 to 50 ppb with a number of occurrences
>50 ppb. The two highest daily maxima were 60 and 62 ppb. The data also show
much lower concentrations during summer, with concentrations typically ranging
between 20 and 30 ppb. Oltmans et al. (2008) examined the time series of O3 and
back trajectories reaching Trinidad Head. They found that springtime maxima (April-
May) were largely associated with back trajectories passing over the Pacific Ocean
and most likely entraining emissions from Asia, with minimal interference from local
sources. However, Parrish et al. (2009) noted that only considering trajectories
coming from a given direction is not sufficient for ruling out local continental
influences, as sea breeze circulations are complex phenomena involving vertical
mixing and entrainment of long-shore components. They found that using a wind
speed threshold in addition to a criterion for wind direction allowed determination of
background trajectories not subject to local influence. This was confirmed by
measurements of chemical tracers of local influence such as CO2, MTBE and radon.
By applying the two criteria for wind speed and direction, they found that Trinidad
Head met these criteria only 43% of the time during spring. Goldstein et al. (2004)
used CO2 as an indicator of exchange with the local continental environment and
found that O3 concentrations were higher by about 2-3 ppb when filtered against
local influence indicating higher O3 in air arriving from over the Pacific Ocean.
At other times of the year, Trinidad Head is less strongly affected by air passing over
Asia and the northern Pacific Ocean; and many trajectories have long residence times
over the semi-tropical and tropical Pacific Ocean where O3 concentrations are much
lower than they are at mid-latitudes. The use of the Trinidad Head data to derive
contributions from background sources requires the use of screening procedures
adopted by Parrish et al. (2009) and the application of photochemical models to
determine the extent either of titration of O3 by fresh NOX emissions and the extent
of local production of O3 from these emissions. Although O3 concentrations at
Trinidad Head might at times be representative of Pacific air arriving over the
3-37
-------
U.S., they cannot be viewed as NA background over the continental U.S. because
of deposition to the surface and chemical loss over the continental United States.
As noted above, anthropogenic emissions from North America also contribute to
hemispheric background and must be filtered out from observations at coastal sites
such as Trinidad Head even when it is thought that air sampled came directly from
over the Pacific Ocean and was not influenced by local pollutant emissions.
Trinidad Head
0.07
0.06
99999999999999999999999999999999
Time (Days: April 19, 2002 - December 31, 2009
Source: Reprinted with permission of Elsevier Ltd., (Oltmans et al.. 2008): and NOAA Climate Monitoring Diagnostics Laboratory for
data from 2008-2009.
Figure 3-8 Time series of MDA8 Os concentrations (ppm) measured at
Trinidad Head, CA, from April 18, 2002 through December 31, 2009.
Parrish et al. (2009) also examined data obtained at other marine boundary layer sites
on the Pacific Coast. These include Olympic NP, Redwood NP, Point Arena, and
Point Reyes. Using data from these sites, they derived trends in O3 of+0.46 ppb/year
(with a 95% confidence interval of 0.13 ppb/year) during spring and +0.34 ppb/year
(0.09 ppb/year) for the annual mean O3 increase in air arriving from over the Pacific
during the past two decades. Although O3 data are available from the Channel
Islands, Parrish et al. (2009) noted that these data are not suitable for determining
background influence because of the likelihood of circulating polluted air from the
South Coast Basin.
3-38
-------
The 2010 Intercontinental Chemical Transport Experiment Ozone Network Study
(IONS-2010) and Research at the Nexus of Air Quality and Climate Change
(CalNex) study conducted in May through June of 2010 had discerning the
contributions of Asian emissions to air quality in California as a major focus. Cooper
et al. (2011) examined O3 profiles measured above four coastal sites in California,
including Trinidad Head. Based on trajectory analyses coupled with comparison with
the O3 profiles, they suggested that Asian pollution, stratospheric intrusions, and
international shipping made substantial contributions to lower tropospheric O3
(typically 0 to 3 km above sea level, meant as a rough approximation of planetary
boundary layer height) measured at inland California sites. These contributions
tended to increase on a relative basis in going from south to north. In particular, no
contribution from local pollution was needed to explain lower tropospheric O3 in the
northern Central Valley; and the contribution of local pollution to lower tropospheric
O3 in the LA basin ranged from 32 to 63% (depending on layer depth; either 0 to
1.5 km or 0 to 3 km). It should be noted that the extent of photochemical production
and loss occurring in the descending air masses between the coastal and inland sites
remains to be determined. Cooper et al. (2011) also note that very little of the O3
observed above California reaches the eastern United States. However, this does not
necessarily mean that the pathways by which Asian O3 could reach the eastern U.S.
were fully captured in this analysis.
Lin et al. (2012) used the AM3 model (-50 x 50 km resolution globally) and satellite
data to characterize the influence of Asian emissions and stratospheric intrusions on
O3 concentrations in southern California and Arizona during CalNex (May-June
2010). The model simulates sharp O3 gradients in the upper troposphere and the
interweaving and mixing of stratospheric air and Asian plumes. Similar phenomena
were also found during field campaigns conducted in the North Atlantic as noted in
Annex AX2.3.1 of the 2006 O3 AQCD (U.S. EPA. 2006b) and introduces
uncertainty into attempts to attribute O3 to these sources, based solely on
observations, because this mixing will affect relationships between CO (mainly a
marker for polluted air that is commonly used to separate air influenced by
anthropogenic pollution from stratospheric air) and O3 (a pollutant and a
stratospheric component). Lin et al. (2012) found that Asian emissions contributed
from 20-30% to O3 in the mid troposphere over the California coast and remnants of
stratospheric intrusions contributed from 50 to 60% to O3 in discrete layers in the
same altitude range. This O3 then has the potential to mix downward into the
planetary boundary layer. Lin et al. (2012) also found evidence of Asian
contributions of up to 8 -15 ppb in surface air during strong transport events in
southern California. These contributions are in addition to contributions from
dominant local pollution sources. Their results suggest that the influence of
background sources on high surface O3 concentrations is not always confined to high
elevation sites. However, it is not clear to what extent the contributions inferred by
Cooper et al. (2011) and Lin et al. (2012) are likely to be found in other years or how
long they would extend into summer.
3-39
-------
3.4.3 Estimating Background Concentrations
Historically, two approaches to estimating NA background concentrations
(previously referred to as PRB) have been considered in past O3 assessments. In the
1996 and earlier O3 AQCDs, measurements from remote monitoring sites were used.
In the 2006 O3 AQCD, the use of CTMs was adopted, because as noted in
Section 3.9 of the 2006 O3 AQCD (U.S. EPA. 2006b). estimates of background
concentrations cannot be obtained directly by examining measurements of O3
obtained at relatively remote monitoring sites in the U.S. because of the long-range
transport from anthropogenic source regions within North America. The 2006 O3
AQCD (U.S. EPA, 2006b) also noted that it is impossible to determine sources of O3
without ancillary data that could be used as tracers of sources or to calculate
photochemical production and loss rates. As further noted by Reid et al. (2008), the
use of monitoring data for estimating background concentrations is essentially
limited to the edges of the domain of interest. The current definition of
NA background implies that only CTMs (see Section 3.3 for description and
associated uncertainties) can be used to estimate the range of background
concentrations. An advantage to using models is that the entire range of O3
concentrations measured in different environments can be used to evaluate model
performance. In this regard, data from the relatively small number of monitoring sites
at which large background contributions are expected are best used to evaluate model
predictions.
Estimates of NA background concentrations in the 2006 O3 AQCD (U.S. EPA.
2006b) were based on output from the GEOS-Chem (v4.3.3) model (Fiore et al..
2003) with 2° x 2.5° horizontal resolution. The GEOS-Chem model estimates
indicated that NA background O3 concentrations in eastern U.S. surface air were
25 ± 10 ppb (or generally 15-35 ppb) from June through August, based on conditions
for 2001. Values reported by Fiore et al. (2003) represent averages from 1 p.m. to 5
p.m.; all subsequent values given for background concentrations refer to MDA8 O3
concentrations. Background concentrations decline from spring to summer.
Background O3 concentrations may be higher, especially at high altitude sites during
the spring, due to enhanced contributions from (1) pollution sources outside North
America; and (2) stratospheric O3 exchange. At the time, only the GEOS-Chem
model (Harvard University. 201 Ob) was documented in the literature for calculating
background O3 concentrations (Fiore et al.. 2003). The simulated monthly mean
concentrations in different quadrants of the U.S. were typically within 5 ppbv of
observations at CASTNET sites, with no discernible bias, except in the Southeast in
summer when the model was 8-12 ppbv too high. This bias was attributed to
excessive background O3 transported in from the Gulf of Mexico and the tropical
Atlantic Ocean in the model (Fiore et al.. 2003).
Although many of the features of the day-to-day variability in O3 at relatively remote
monitoring sites in the U.S. were simulated reasonably well by GEOS-Chem (Fiore
et al.. 2003). uncertainties in the calculation of the temporal variability of O3
originating from different sources on shorter time scales must be recognized.
The uncertainties stem in part from an underestimate in the seasonal variability in the
3-40
-------
STE of O3 (Fusco and Logan, 2003), the geographical variability of this exchange,
and the variability in the exchange between the free troposphere and the PEL in the
model. In addition, the relatively coarse spatial resolution in that version of GEOS-
Chem (2° x 2.5°) limited the ability to provide separate estimates for cities located
close to each other, and so only regional estimates were provided for the 2006 O3
AQCD (U.S. EPA. 2006b) based on the results ofFioreetal. (2003).
Wang et al. (2009a) recomputed NA background concentrations for 2001 using
GEOS-Chem (V7-01-01) at higher spatial resolution (1° x 1°) over North America
and not only for afternoon hours but for the daily maximum 8-h avg O3
concentration. The resulting background concentrations, 26.3 ± 8.3 ppb for summer,
are consistent with those of 26 ± 7 ppb for summer reported by Fiore et al. (2003),
suggesting horizontal resolution was not a substantial factor limiting the accuracy of
the earlier results. In addition to computing NA background concentrations, Wang et
al. (2009a) also computed U.S. background concentrations of 29.6 ± 8.3 ppb with
higher concentrations in the Northeast (up to 15 ppb higher) and the Southwest (up to
13 ppb higher) for summer means.
3.4.3.1 Updated GEOS-Chem Model Estimates of Background
Concentrations
Zhang et al. (2011) computed NA background, U.S. background and natural
background O3 concentrations using GEOS-Chem (v8-02-03) at an even finer grid
spacing of 0.5° x 0.667° over North America for 2006 through 2008. For March
through August 2006, mean NA background O3 concentrations of 29 ± 8 ppb at low
elevation (<1,500 meters) and 40 ± 8 ppb at high elevation (>1,500 meters) were
predicted. Spring and summer mean O3 concentrations calculated for the base case
(i.e., including all natural and anthropogenic sources worldwide), U.S. background,
and NA background in surface air for spring and summer 2006 calculated by Zhang
et al. (2011) are shown in the upper, middle and lower panels of Figure 3-9.
3-41
-------
Base: Spring (49 ppbv)
Base: Summer (52 ppbv)
-130 -120 -111) -1(1(1 -!)() -80 -70
Longitude (degrees)
-130 -120 -HI) -Kill -00 -MO -70
Longitude (degrees)
USB: Spring (36 ppbv)
USB: Summer (33 ppbv)
-130 -120 -110 -11)0 -90 -80 -7
Longitude (degrees)
-130 -120 -110 -Kill -90 -80 -70
Longitude (degrees)
NAB: Spring (33 ppbv)
NAB: Summer (30 ppbv)
-130 -120 -110 -100 -90 -80 -70
Longitude (degrees)
10 -90 -80 -70
Longitude (degrees)
15
25
35
45
55
Note: Values in parentheses refer to continental U.S. means and are shown as black squares in the color bar for summer and white
squares for spring.
Source: Adapted from Zhang et al. (2011).
Figure 3-9 Mean MDA8 Os concentrations in surface air for spring and
summer 2006 calculated by GEOS-Chem for the base case (Base),
U.S. background (USB), and NA background (NAB).
3-42
-------
As noted above, Zhang et al. (2011) found increases in Asian emissions only
accounted for an average increase of between 1 to 2 ppb in background O3 across the
U.S. even though Asian emissions have increased by about 44% from 2001 to 2006.
As can be seen from Figure 3-9, U.S. background and NA background
concentrations are very similar throughout most of the United States. Zhang et al.
(2011) also found that NA background concentrations are ~4 ppb higher, on average,
in the 0.5° x 0.667° version than in the coarser 2° x 2.5° version. This difference was
partially due to higher resolution (~1 to 2 ppb) and the remainder to the combination
of changes in lightning and Asian emission estimates as well as higher model
resolution.
As can be seen from the middle and lower panels in Figure 3-9, U.S. background and
NA background concentrations tend to be higher in the West, particularly in the
Intermountain West and in the Southwest compared to the East in both spring and
summer. U.S. background and NA background concentrations tend to be highest in
the Southwest during summer in the GEOS-Chem model, driven by lightning NOX.
Intercontinental transport and stratospheric intrusions are major contributors to the
high elevation, Intermountain West during spring with wildfires becoming more
important sources during summer. The base case O3 concentrations (upper panels)
show two broad maxima with highest concentrations extending throughout the
Southwest, Intermountain West, and the East; in both spring and summer. These
maxima extend over many thousands of kilometers demonstrating that O3 is a
regional pollutant. Low-level outflow from the Northeast out over the Atlantic Ocean
and from the Southeast out over the Gulf of Mexico is also apparent.
Note that background concentrations tend to increase with increasing base model
(and measured) concentrations at higher elevation sites, particularly during spring.
However, it is not only background O3 that increases with elevation. Convection
efficiently lifts precursor emissions and O3 from the polluted boundary layer in
North America as well as in Europe and Asia to the mid and upper troposphere.
Significant production of O3 from precursor emissions and from lightning can occur
aloft, providing a diffuse source in the mid and upper troposphere because the O3
production efficiency with respect to NOX is much higher aloft than at low
elevations. Higher wind speeds aloft, coupled with a lifetime of O3 that increases
with height in the troposphere, allow pollutants to be transported much farther and to
mix over wider areas than they would at lower elevations. Thus, through these
mechanisms, O3 aloft at a given location can be higher than at low elevations and
caution should be observed in attempting to ascribe increases in O3 with elevation to
particular sources.
Although the results of Zhang et al. (2011) are broadly consistent with results from
earlier coarser resolution versions of GEOS-Chem used by Fiore et al. (2003) and
Wang et al. (2009a), there are some apparent differences. Concentrations of O3 for
both the base case and the NA background case in Zhang et al. (2011) are higher in
the Intermountain West than in earlier versions. In addition, background
concentrations in many eastern areas tend to be higher on days when predicted total
3-43
-------
O3 is > 60 ppb or at least do not decrease with increasing total O3 Zhang et al.
(2011).
Figure 3-10 shows seasonal mean estimates of contributions to O3 from Canadian
and Mexican emissions calculated by Zhang et al. (2011) as the difference between
U.S. background and NA background values and then averaged over spring and
summer following the procedure in Wang et al. (2009a). U.S. background
concentrations are on average 3 ppb higher than NA background concentrations
during spring and summer across the United States. Highest values in Figure 3-10 (in
the U.S.) are found over the Northern Tier of New York State (19.1 ppb higher than
NA background) in summer. High values are also found in other areas bordering
Canada and Mexico. Although the contributions from Canada and Mexico were
obtained by differencing, it should be remembered that relations between O3 and
precursors are subject to non-linear effects that are strongest near concentrated
sources of precursors, as noted in Section 3.2.4. Therefore, the values shown in the
figure are only estimates of contributions to total O3 coming from Canada and
Mexico.
CM: Spring (3 ppbv)
CM: Summer (3 ppbv)
-130 -120 -111) -UK) -90 -SO -TO
Longitude (degrees)
-130 -120 -110 -100 -90 -80
Longitude (degrees)
o
10
Note: Values in parentheses show mean difference (ppb) across the U.S.
Source: Adapted from Zhang et al. (2011).
Figure 3-10 Spring and summer mean Canadian and Mexican (CM)
contributions to MDA8 O3 determined as the difference between the
U.S. background and NA background.
Figure 3-11 shows MDA8 O3 concentrations for spring (March-May) and summer
(June-August) 2006 simulated by GEOS-Chem (versus measured) by the ensemble
of CASTNET sites in the Intermountain West, Northeast, Great Lakes, and Southeast
(Zhang et al.. 2011). Shown is the 1:1 line and NA background (blue) and natural
background (green) model statistics as box plots (minimum, 25th, 50th, 75th
3-44
-------
percentile, and maximum) for 10-ppbv bins of observed O3 concentrations. These
plots show that NA background constitute a larger fraction of modeled base case O3
at the upper end of the concentration distribution for the Intermountain West than for
other regions of the country.
Spring
Summer
100
eo
GO
40
20
0
100
90
60
dQ
Q.
°r 20
Ł o
V 100
l/l
2 so
^ 60
40
20
0
100
BO
eo
40
20
0
- Northeast
Intermountain West
- Great Lakes
Southeast
- HA backgrcxgnd
: Natural
0 20 40 60 80 100 0 20 40 60 80 100
Observed Ozone [ppbv]
Note: Shown is the 1:1 line and North American (NA) background (blue) and natural background (green) model statistics as box
plots (minimum, 25th, 50th, 75th percentile, and maximum) for 10-ppbv bins of observed O3 concentrations.
Source: Adapted from Zhang et al. (2011).
Figure 3-11 MDA8 O3 concentrations for spring (March-May) and summer
(June-August) 2006 simulated by GEOS-Chem vs. measured by the
ensemble of CABINET sites in the Intermountain West, Northeast,
Great Lakes, and Southeast.
3-45
-------
Comparisons between GEOS-Chem and measurements of the mean MDA8 O3
between March and August at individual CASTNET sites across the country are
shown as supplemental material in Section 3.8, Figure 3-58 through Figure 3-64.
In general, the GEOS-Chem predictions tend to show better agreement with
observations at the high-altitude sites than at the low-altitude sites. Overall
agreement between model results for the base case and measurements is within a few
parts per billion for spring-summer means in the Northeast (see Figure 3-58 in
Section 3.8) and the Southeast (see Figure 3-59 in Section 3.8). except in and around
Florida where the base case over-predicts O3 by 10 ppb on average. In the Upper
Midwest (Figure 3-60 in Section 3.8). the Intermountain West (Figure 3-61 and
Figure 3-62 in Section 3.8). and the West (Figure 3-63 in Section 3.8) including most
sites in California (Figure 3-64 in Section 3.8). the model predictions are within
5 ppb of measurements. The model under-predicts O3 by 10 ppb at the Yosemite site
(Figure 3-64 in Section 3.8). These results suggest that the model is capable of
calculating March to August mean MDA8 O3 to within ~5 ppb at most (26 out of 28)
sites chosen.
Comparison between results in Wang et al. (2009a) for 2001 with data obtained in the
Virgin Islands indicate that GEOS-Chem over-predicts summer mean MDA8 O3
concentrations there by 10 ppb (28 vsersus 18 ppb). Ozone concentrations at the
Virgin Islands NP site appear not to have been affected by U.S. emissions, based on
the close agreement between the base case and the NA background case. Wind roses
calculated for the Virgin Islands site indicate that wind patterns affecting this site are
predominantly easterly/southeasterly in spring and summer. The over-predictions at
the Virgin Islands site imply that modeled O3 over the tropical Atlantic Ocean is too
high. As a result, inflow of O3 over Florida and into the Gulf of Mexico is also likely
to be too high as winds are predominantly easterly at these low latitudes. Similar
considerations apply to the results of Zhang et al. (2011). Possible explanations
include deficits in model chemistry (for example, reactions involving halogens are
not included) and/or subsidence that is too strong over tropical oceans in the model.
No clear explanation can be provided on why the model under-predicts mean O3 at
Yosemite (elevation 1,680 meters) by ~10 ppb (see Figure 3-64 in Section 3.8).
However, March to August mean MDA8 O3 concentrations are simulated to within a
few parts per billion at an even higher elevation site in California (Converse Station,
elevation 1,837 meters) and at the low elevation sites.
Figure 3-65 in Section 3.8 shows a comparison of GEOS-Chem output with
measurements at Mt. Bachelor, OR and Trinidad Head, CA from March-August,
2006 from Zhang et al. (2011). For the Mt. Bachelor model runs, model estimates are
given for both a coarse (2° x 2.5°) and fine (0.5° x 0.667°) resolution model.
In general, mean concentrations are simulated reasonably well at both coarse and
finer grid resolution versions of the model with mean values 2 ppb higher in the finer
resolution model. Although the finer resolution version provides some additional day
to day variability and can capture the timing of peaks, it still does not adequately
resolve peak concentrations as can be seen for an event in the second half of April.
3-46
-------
Figure 3-66 in Section 3.8 shows a comparison of vertical profiles (mean ± lo)
calculated by GEOS-Chem with ozonesondes launched at Trinidad Head, CA and
Boulder, Colorado. As can be seen from the figure, variability in both model and
measurements increases with altitude, but variability in the model results is much
smaller at high altitudes than seen in the observations. This may be due in part to the
inability of grid-point models to capture the fine-scale, layered structure often seen in
O3 in the mid and upper troposphere (Rastigeiev et al.. 2010: Newell et al.. 1999) and
to inadequacies in parameterizations of relevant chemistry and dynamics.
Figure 3-67 and Figure 3-68 in Section 3.8 show a comparison of vertical profiles
simulated by AM3 at 50 x 50 km global resolution (Lin et al.. 2012) with
ozonesondes launched at several locations in California during May-June 2010. Note
that in contrast to comparing measured mean monthly O3 profiles to monthly mean
profiles calculated by GEOS-CHEM (see, for example, Figure 3-66 in Section 3.8).
AM3 is sampled for comparison to individual measurements of O3 profiles. This
model has likely had the most success in simulating vertical O3 gradients in the
upper troposphere and in capturing layered structures in the mid and upper
troposphere.
The natural background for O3 averages 18 ± 6 ppbv at the low elevation sites and
27 ± 6 ppbv at the high elevation sites in the GEOS-Chem model Zhang et al. (2011).
In regions where non-linear effects are small, far from concentrated sources of O3
precursors, the difference between NA background and natural background O3
concentrations provides an estimate of contributions from intercontinental pollution
including anthropogenic methane (given by the difference between values in 2006
and the pre-industrial era, or 1,760 ppb and 700 ppb). The difference between the
two backgrounds averages 9 ppbv at the low elevation sites and 13 ppbv at sites in
the Intermountain West. Based on the Zhang et al. (2011) model runs, anthropogenic
methane emissions are estimated to contribute ~4-5 ppb to global annual mean O3
surface concentrations. North American emissions of methane are uncertain, but are a
small fraction of total anthropogenic input. This suggests that slightly less than half
of the difference between North American background and natural background is due
to the increase of methane since the beginning of the industrial era and the other half
is due to anthropogenic emissions of shorter lived VOCs and NOX. However, the
relative importance of methane for O3 production is expected to increase in the
future. Indeed, variations in methane concentrations account for approximately 75%
of the wide spread (~5 ppb) in tropospheric O3 projections between Representative
Concentration Pathway (RCP) scenarios for the next century (Wild et al.. 2012): see
Section 10.3.6.1 in Chapter 10 for more on the RCP scenarios.
Figure 3-12 shows frequency distributions for observations at low-altitude and high-
altitude CASTNET sites along with GEOS-Chem frequency distributions for the base
case, NA background, and natural background. Most notable is the shift to higher
concentrations and the narrowing of the concentration distributions for all three
simulations and the observations in going from low to high altitudes. However,
maximum concentrations show little if any dependence on altitude, except for the
natural background (which tends to be slightly higher at high altitude sites).
3-47
-------
0.10
0.05
0.00
0.10
0.05
0.00
Low-altitude sites (< 1.5 km)
observation
GEOS-Chem
NA background -
Natural
'18
High-altitude sites (>1.5 km)
0
18
I
20
40 60
Ozone [ppbv]
80
-Jo
100
Note: Observations (black) as well as GEOS-Chem estimates for the base case (red), NA background (blue), and natural
background (green dashed).
Source: Zhang et al. (2011).
Figure 3-12 Frequency distributions of MDA8 Os concentrations in March-
August 2006 for the ensemble of low-altitude (<1,500 meters) and
high-altitude CASTNET sites (>1,500 meters) in the U.S.
3.4.3.2 Using Other Models to Estimate Background
Concentrations
Another approach to modeling background concentrations involves using a regional
CTM such as CMAQ or CAMx with boundary conditions taken from a global scale
CTM such as GEOS-Chem (see Section 3.3 for discussion of this approach). Mueller
and Mallard (201 la), while not calculating NA background values exactly as defined
here, calculated contributions from natural sources and inflow from the boundaries to
O3 for 2002 using MM5 and CMAQ for the outermost domain (36 km resolution)
shown in Figure 3-4 with boundary conditions from GEOS-Chem. The overall bias
based on comparison with AQS monitors for the base case is about 3 ppb; the annual
mean fractional bias and mean fractional error were 7% and 21% for the ozone
season across the United States. Note that Figure 2 in their paper is mislabeled, as it
3-48
-------
should refer to the case with total emissions - not to natural emissions in North
America only (Mueller and Mallard, 201 Ib). However, boundary conditions are
fixed according to monthly averages based on an earlier version of GEOS-Chem and
do not reflect shorter term variability or trends in Northern Hemispheric emissions of
pollution. In addition, fluxes of O3 from the stratosphere are not included explicitly.
Note that their natural background includes North American natural background
emissions only and influence from boundary conditions and thus is not a global
natural background. Calculated values including natural emissions from North
America and from fluxes through the boundaries are somewhat larger than given in
Zhang et al. (2011). in large measure because of much larger contributions from
wildfires and lightning. Wildfire contributions reach values of-140 ppb in
Redwoods National Park, CA and higher elsewhere in the U.S. and in Quebec in the
simulations by Mueller and Mallard (201 la). Lightning contributions (ranging up to
-30 ppb) are substantially larger than estimated by Kaynak et al. (2008) (see
Section 3.4.1.2). The reasons for much larger contributions from wildfires and
lightning found by Mueller and Mallard (2011 a) are not clear and need to be
investigated further.
Emery et al. (2012) used CAMx with boundary conditions taken from the coarse
resolution version of GEOS-Chem (2 x 2.5° or -200 km resolution) to derive
NA background concentrations of O3. The nested CAMx simulations were run at a
horizontal resolution of 12 km separately for the eastern and western United States.
The following paragraphs compare results from the Emery et al. (2012) nested
CAMx simulations at 12 km resolution, with those obtained by Zhang et al. (2011)
using GEOS-Chem simulations at 0.5° x 0.667° (-50 km) resolution. This is in
contrast to the comparison reported in Emery et al. (2012) in which results from
CAMx at 12 km resolution were compared to results from the 2° x 2.5° (-200 km)
resolution version of the GEOS-Chem model over the United States.
Figure 3-13 shows seasonal mean MDA8 O3 concentrations calculated by Emery et
al. (2012) using CAMx for 2006 for the base case and for NA background.
Figure 3-14 shows a comparison of monthly average O3 concentrations calculated by
GEOS-Chem (Zhang et al., 2011) with those calculated by CAMx (Emery et al.,
2012). Comparison of the base case for GEOS-Chem with that for CAMx in
Figure 3-14 indicates broad agreement in spatial patterns.
3-49
-------
Base: Spring (45 ppbv)
Base: Summer (52 ppbv)
Longitude (degrees)
Longitude (degrees)
NAB: Spring (32 ppbv)
Longitude (degrees)
NAB: Summer (33 ppbv)
Longitude (degrees)
15
25
35
45
55
65
75
Note: Values in parentheses refer to continental U.S. means and are shown as black squares in the color bar for summer and white
squares for spring.
Source: Adapted from Emery et al. (2012).
Figure 3-13 Mean MDA8 Oz concentrations in surface air during spring and
summer 2006 (top) calculated by GEOS-Chem/CAMx for the base
case (Base, top) and NA background (NAB, bottom).
3-50
-------
Sites above 1.5km
Month in 2006
Sites below 1.5km
Month in 2006
Note: Shaded area shows 1 SD range about the mean of observations.
Source: Adapted [with permission of Elsevier, Emery et al. (2012)1 and from Zhang et al. (2011).
Figure 3-14 Monthly average MDA8 Os concentrations observed (Obs) and
predicted for the base case and NA background (NAB) by GEOS-
Chem (GC) and GEOS-Chem/CAMx (CX) at CASTNET sites above
1,500 meters elevation (upper panel) and CASTNET sites below
1,500 meters elevation (lower panel).
3-51
-------
Supplemental figures (Figure 3-69 through Figure 3-74) in Section 3.8 show box
plots comparing MDA8 O3 concentrations calculated by GEOS-Chem at
0.5° x 0.667° resolution and CAMx for March-August 2006 at the combined set of
CASTNET sites used by both groups for model evaluation. Note that the individual
model results and the observations are unpaired in time. At CASTNET sites in the
Northern Rockies, both models tend to under-predict maximum O3 concentrations,
but they are generally higher in CAMx than in GEOS-Chem (typically by 5-10 ppb).
The distribution of MDA8 values from GEOS-Chem is consistent with measured
distributions (i.e., cannot be rejected using Mann-Whitney rank sum test, p-value
<0.01) at 18 of 39 sites in spring and 21 of 39 sites in summer. The distribution of
MDA8 values from CAMx is consistent with measured distributions at 13 of 39 sites
in spring and 18 of 39 sites in summer. When spring and summer are pooled, both
simulations are consistent with measured distributions at 16 out of 39 sites (but not
the same 16 sites). There are examples in which either model over- or under-
simulates maximum concentrations. However, over-predictions are made more often
by CAMx. At high elevations in the Intermountain West (see Figure 3-72 in
Section 3.8). both models tend to under-predict maxima, but their interquartile range
agrees much better with observations. As McDonald-Buller et al. (2011) noted,
complex topography in some regions of the U.S. could influence surface O3 through
fine-scale, orographically induced flow regimes. In addition, numerical diffusion
broadly affects the ability of models to capture observed maxima, particularly at
mountain sites. McDonald-Buller et al. (2011) also note there are regions in the U.S.
where global models show consistent biases. For example, models are generally
unable to simulate the low O3 concentrations observed at Gulf Coast sites in summer
during onshore flow from the Gulf of Mexico, which could reflect marine boundary
layer chemistry and/or stratification that is not properly represented in the model.
Both models over-predict O3 at two sites in Florida (Sumatra and Indian River
Lagoon). However, further inland, CAMx tended to over-predict O3 at the
Coffeeville, MS; Sand Mountain, AL; and Georgia Station, GA sites whereas GEOS-
Chem did not. The same is true for higher elevation sites (Great Smoky Mountain,
NC-TN; Shenandoah, VA). In the Northeast, there is a general tendency for both
models to over-predict the measured distributions with somewhat higher maximum
concentrations in CAMx compared to GEOS-Chem and observations (see
Figure 3-69 to Figure 3-74 in Section 3.8).
The most readily discernible differences in model formulation are in the model grid
spacing and the treatment of wildfires. The finer resolution in CAMx allows for
topography to be better-resolved producing higher maximum O3 concentrations in
the Intermountain West. For wildfires, treatment differences include emission
composition, emission time averaging, and associated chemistry. Wildfires produce
more O3 in CAMx simulations than in GEOS-Chem simulations, and Emery et al.
(2012) attribute these enhancements to shorter emission time averaging. The CAMx
emissions average fire emissions at hourly resolution based on the SmartFire
algorithm, whereas GEOS-Chem uses monthly averages from GFED2. Each model
representation also uses different emission compositions. The emissions used by
Emery et al. (2012) include a larger number of VOCs and additional categories of
VOCs than used by Zhang et al. (2011). Following emission, Emery et al. (2012)
3-52
-------
note that photochemical aging of wildfire emissions depends on the chemical
mechanism. Neither chemical mechanism was designed specifically for these type of
events. GEOS-Chem has traditionally focused on the chemistry of the non-urban
troposphere and does not represent secondary products of fast reacting VOCs as does
CB05. A lack of reactivity of secondary products would cause a dampening of fire
contributions to O3. CB05 has traditionally focused on urban chemistry and does not
explicitly includes ketones (Henderson et al.. 2011). which are among the top ten
VOCs emitted from fires (Andreae and Merlet 2001). The O3 increases seen in
Emery et al. (2012) and Mueller and Mallard (2011 a), however, are subject to
uncertainties in the representation of physics in the wildfire plumes.
The improvements in characterizing emissions would lead to smoke plumes that
attenuate light, thereby reducing photolysis and photoreactivity (e.g.. Real et al..
2007). The wildfires would also alter temperature and convective activity that
influences plume rise and the height of the planetary boundary layer. Emery et al.
(2012) note the need for more research to improve simulation of O3 from fires. Using
a sensitivity analysis of CAMx, the authors showed that removing wildfires in the
West (California, Oregon, and Idaho) resulted in reductions of NA background O3 of
10 to 50 ppb, with smaller reductions elsewhere. Further, Emery et al. (2012) note
that their calculated O3 increases in the vicinity of wildfires is consistent with that of
Mueller and Mallard (2011 a).
Emery et al. (2012) captured the timing of a possible stratospheric intrusion at
Gothic, Colorado, on April 19-20, 2006, and predicted an MDA8 O3 value of
~73 ppb using CAMx on April 20, compared to a measured observation value of
87 ppb. GEOS-Chem (at 0.5° x 0.667°) predicted -65 ppb for this event. The higher
spatial resolution in CAMx likely contributed to the improvement in model
performance, but this may not be the only factor. AM3, another global scale CTM
(Lin et al.. 2012) at ~2° x 2.5° resolution predicted ~75 ppb for that event; suggesting
that differences in dynamical cores between WRF and AM3, different treatments of
the stratospheric O3 source, and perhaps the spatial extent of the intrusion's effect on
surface O3 should be considered in addition to model resolution. Typically, these
strong intrusions have spatial extent of thousands of kilometers along their axis and
hundreds of kilometers transversely to this axis, as noted in the 2006 O3 AQCD (U.S.
EPA. 2006b). Note that all three models (CAMx, GEOS-Chem, and AM3) under-
predicted the magnitude of this event. These results indicate a need for process-
oriented evaluation and targeted measurements that yield insight into both chemical
and dynamical processes. The R2 for comparison of AM3 with observations of
MDA8 O3 from March-August 2006 was 0.33, with lower R2 for GEOS-Chem and
CAMx. All three models predicted very similar means for March to August 2006:
54.9 ppb (simulated by the fine resolution version of GEOS-Chem), 55.0 ppb
(simulated by CAMx), and 58.6 ppb (simulated by AM3), compared to 55.9 ppb for
observed concentrations. See Figure 3-75 in Section 3.8: however, note that
Figure 3-75 does not show the CAMx mean of 55.0 ppb, which was reported in
Emery et al. (2012).
The results from either model have also been compared to more urban oriented sites
in the AQS network. As noted earlier, comparisons between model results and
3-53
-------
observations become problematic near concentrated sources of O3 precursors (NOX
and VOCs) in urban cores. Emery et al. (2012) note that in coarse resolution models
rural biogenic and urban precursor emissions are mixed immediately leading to
higher production efficiency for O3. Finer resolution models are better able to
separate these two source categories and to resolve features of urban chemistry such
as titration of O3 by NOX emitted by traffic and subsequent processing of NOX
emissions during transport downwind. CAMx at 12 x 12 km resolution is better able
to capture these features than GEOS-Chem at 50 x 50 km resolution. Both models
tend to over-predict O3 at the low O3 concentrations in areas where O3 scavenging
by NOX is evident. In these situations, NA background O3 concentrations are often
higher than in the respective models for the base case. At high O3 concentrations
downwind of source areas, both models predict NA background O3 concentrations
that are much lower than observed or base case O3. The latter results are in accord
with results shown in Figure 3-11 for rural CASTNET sites at low elevations, which
show lower ratios between NA background O3 and either observations or base case
O3 at high O3 than at low O3 concentrations.
Figure 3-15 shows the annual 4th-highest MDA8 O3 predicted by GEOS-Chem (at
0.5° x 0.667° resolution) for the base case (upper panel), and corresponding
U.S. background (middle panel) and NA background (lower panel) MDA8 O3 on the
same days for 2006. Figure 3-16 shows corresponding values predicted by CAMx for
the base case (upper panel) and NA background (lower panel) MDA8 O3 on the
same days for 2006. As can be seen from Figure 3-15 and Figure 3-16, on those days
when models predicted their annual 4th-highest MDA8 O3, the corresponding
NA background concentrations are 36 ± 9 ppb in the eastern United States. Base case
concentrations are much higher indicating that regional pollution is mainly
responsible for the models 4th-highest concentrations. In the western U.S. on the
other hand, NA background concentrations are generally higher and make up a larger
fraction of the calculated 4th-highest MDA8 O3 in both models, but for different
reasons. GEOS-Chem predicts highest values in the Southern Rockies because of
over-production of NOX by lightning. CAMx predicts highest values in ID, OR and
WA from wildfires. The CAMx run includes day specific values for area burned, but
GEOS-Chem uses monthly averages. (A more recent version of GEOS-Chem also
incorporates day specific estimates for area burned.) Remaining areas of relatively
high background levels (>60 ppb) are due mainly to some combination of
stratospheric intrusions and Eurasian emissions. There are a few examples that can
be used to give a rough idea of the magnitudes of episodically high background
contributions. A comparison of the annual 4th-highest MDA8 O3 concentration
simulated by CAMx including wildfires and omitting them indicates that wildfires
contributed ~ 30 to 40 ppb in Idaho, Montana, and Washington with a potentially
larger contribution in the upper northwestern corner of California. Estimated
contributions from strong stratospheric intrusions to surface O3 in AM3 could range
up to ~ 55 ppb in the western United States. It should be borne in mind in using these
figures, that they are model derived and hence could be model specific. Issues related
to calculating O3 formation in wild fire plumes by CAMx were mentioned above.
The method for calculating the stratospheric contributions in AM3 is based on the
amount of O3 carried downward through a given surface thus raising the possibility
3-54
-------
that O3 could be generated in the upper troposphere, transported into the stratosphere
and then transported downward again. This could indeed lead to an overestimate of
the baseline stratospheric attribution in the model. However, Lin et al. (2012) focus
on the "delta during events" rather than the baseline, because during specific
intrusion events this recycling of O3 between troposphere and stratosphere is less
likely to be a problem. Therefore, there is greater confidence in the anomaly from
stratospheric O3 rather than the absolute amount. Tagging O3 according to where it
was produced will remove this ambiguity in future model runs.
All models undergo continuous updating of inputs, parameterizations of physical and
chemical processes, and improvements in model resolution. Inputs that might be
considered most relevant include emissions inventories, chemical reactions, and
meteorological fields. This leads to uncertainty in model predictions in part because
there is typically a lag between updated information for the above inputs—as
outlined in Section 3.2 for chemical processes and emissions and in Section 3.3 for
model construction—and their implementation in CTMs including GEOS-Chem or
the other models described above. Quantitative estimates of uncertainties from
meteorological and emission inputs and chemical mechanisms are problematic
because simulations designed to quantify uncertainties from these sources have not
been performed for these model runs. At best, these uncertainties can be estimated by
comparison with observations while recognizing that compensating errors likely
exist.
3-55
-------
Base: All Year
20
-130 -120 -lilt -100 -911 -SO
Longitude (degrees)
USB: All Year
_i;!0 _|20 -111) -KX) -30 -80
Longitude (degrees)
NAB: All Year
2(1
-70
— 130 -120 -111) -100 -90 -80 -70
Longitude (degrees)
35
45 55 65 75
jppbv
85 95
Source: Adapted from Zhang et al. (2011).
Figure 3-15 Annual 4th-highest MDA8 O3 predicted by GEOS-Chem
(0.5° x 0.667°) for the base case (Base) with corresponding
U.S. background (USB) and NA background (NAB) MDA8 O3 for the
same days in 2006.
3-56
-------
Base: All Year
35
Longitude (degrees)
NAB: All Year
Longitude (degrees)
45
55
65
75
85
Source: Adapted from Emery et al. (2012).
Figure 3-16 Annual 4th-highest MDA8 Os predicted by CAMx for the base case
(Base) and corresponding NA background (NAB) MDA8 Os for the
same days in 2006.
3-57
-------
Since NA background is a construct that cannot be directly measured, the range of
background O3 concentrations must be estimated using CTMs. Results from the
Zhang etal. (2011) GEOS-Chem and Emery et al. (2012) CAMx model estimates
were chosen for further analysis because these models have produced the latest
estimates for background O3 concentrations documented in the open literature.
The main results from these two modeling efforts can be described as follows:
• Both models show background concentrations vary spatially and temporally;
• Simulated mean background concentrations are highest in the Intermountain
West (i.e., at high altitude) in spring and lowest in the Northeast during
summer;
• Background concentrations tend to increase with total (i.e., base case) O3
concentrations at high elevation; but that tendency is not as clear at low
elevations.
The most pronounced differences between the Zhang et al. (2011) GEOS-Chem and
the Emery et al. (2012) CAMx models—when compared with observations—are at
the upper end of the concentration distribution. At high elevations, differences are
likely to be the result of under-predictions of background contributions which are
driven mainly by episodic events such as stratospheric intrusions and wildfires.
In general, CAMx predicts higher values at the upper end of the concentration
distribution than does GEOS-Chem. At low elevations (<1,500 meters)—located
mainly in the East—the reasons for under-predictions at the upper end of the
concentration distribution are more complex and likely involve extensive interactions
between anthropogenic and natural sources.
Table 3-1 summarizes modeling results for seasonal mean MDA8 O3 by region
simulated by the two models. The regions in Table 3-1 are shown in Figure 3-50.
As can be seen from the table, seasonal means predicted by GEOS-Chem are within
a few parts per billion of measurements in both spring and summer for all regions
shown except for California in the spring. Although CASTNET sites are meant to
represent regional background air, they can be heavily influenced by polluted air
masses, particularly in California where the under-predictions are largest. Seasonal
means are simulated by CAMx to within 2-5 ppb except in California in the spring
where they are under-predicted by 8 ppb and at sites in the Northeast and Southeast
where they are over-predicted by 8-9 ppb in summer. When compared to
observations, the mean R2 within each region—except for California in the spring—
is higher for CAMx than for GEOS-Chem suggesting better ability to track day-to-
day variability by CAMx. It is clear from these results that model resolution (at least
for the model resolutions considered here) is not the dominant factor determining
agreement of the means between simulations or between simulations and
measurements. Differences in model chemistry and physics must also be considered.
Table 3-2 summarizes modeling results for the annual 4th-highest (99th -percentile)
MDA8 O3 for the same seasons and regions used in Table 3-1. As can be seen in the
table, the GEOS-Chem and the CAMx models both underestimate mean paired (day-
specific) 4th-highest values in California by -20 ppb. In general, CAMx simulates
3-58
-------
paired annual 4th-highest MDA8 O3 concentrations that are higher and in better
agreement with measurements. Shown alongside the model estimates is the number
of days the modeled MDA8 O3 concentrations were within 5 ppb of observed
concentrations. The unpaired entries for the models in Table 3-2 show model
predicted 4th-highest MDA8 O3 concentrations that are not calculated on the same
day as the 4th-highest values observed at CASTNET sites. It can be seen that
simulated regional means of the 4th-highest MDA8 O3 are in better agreement with
measurements when results are unpaired by date. In other words, the models do not
necessarily predict the annual 4th-highest MDA8 O3 concentrations on the same day
as they are observed.
These results underscore the uncertainties inherent in any model's attempts to
simulate day specific 4th-highest O3 concentrations. As noted earlier, uncertainties in
calculating day specific O3 concentrations are especially challenging because of the
lack of day specific data for emissions of many species. While progress is being
made in obtaining day specific data for lightning strikes and area burned in wildfires,
the emission factors for precursors from these episodic sources such as lightning and
wildfires are still uncertain. In addition to uncertainty in emissions, uncertainties in
the treatment of transport and chemical mechanisms in the models must also be
considered.
3-59
-------
Table 3-1
Region
California
(5 sites)
West
(14 sites)
North Central
(6 sites)
Northeast
(5 sites)
Southeast
(9 sites)
Comparison of Zhang et al. (2011) and Emery et al. (2012) results
for MDA8 O3 concentrations (ppbv) with measurements at selected
CASTNET sites.
Observation Model
CASTNET0
GEOS-Chemd
Based
U.S. bkge
NA bkgf
CAMxd
Based
NA bkgf
CASTNET0
GEOS-Chemd
Based
U.S. bkge
NA bkgf
CAMxd
Based
NA bkgf
CASTNET0
GEOS-Chemd
Based
U.S. bkge
NA bkgf
CAMxd
Based
NA bkgf
CASTNET0
GEOS-Chemd
Based
U.S. bkge
NA bkgf
CAMxd
Based
NA bkgf
CASTNET0
GEOS-Chemd
Based
U.S. bkge
NA bkgf
CAMxd
Based
NA bkgf
Mean MDA8 O3
concentration (ppbv)a
Spring
58 ± 12°
52 ± 11
38 ±7
37 ±6
50 ± 10
39 ±6
54 ±9°
53 ±7
42 ±6
41 ±6
49 ±8
40 ±7
47 ±10°
47 ±8
33 ±6
30 ±7
45 ± 11
30 ±6
48 ±10°
45 ±7
33 ±7
29 ±6
46 ±11
30 ±5
52 ± 11 °
51 ±7
32 ±7
29 ±7
54 ±9
33 ±6
Summer
69 ± 1 4°
66 ± 18
37 ±9
35 ±9
66 ± 13
42 ±6
55 ±11°
55 ±11
40 ±9
38 ±9
57 ± 10
41 ±8
50 ± 1 2°
51 ±14
27 ±7
24 ±7
54 ± 13
31 ±5
45 ± 1 4°
45 ± 13
24 ±7
18±6
53 ±14
27 ±6
52 ± 1 6°
54 ±9
29 ± 10
28 ±9
61 ± 12
30 ±6
Mean R2
Spring13 Summer13
0.52 0.22
0.50 0.30
0.30 0.12
0.39 0.33
0.52 0.44
0.63 0.48
0.44 0.47
0.53 0.54
0.42 0.21
0.56 0.45
"Seasonal (spring, summer) mean MDA8 O3 concentration ± standard deviation;
bMean R2 of all individual model-measurement pairs at individual CASTNET sites;
°Observed concentrations at CASTNET sites;
dModeled concentrations at CASTNET sites;
eModeled U.S. background concentrations at CASTNET sites,
'Modeled NA background concentrations at CASTNET sites;
Source: Data from Zhang et al. (2011) for GEOS-Chem and Emery et al. (2012) for CAMx.
3-60
-------
Table 3-2 Comparison of annual 4th-highest MDA8 O3 concentrations
measured at CASTNET sites in 2006 with MDA8 O3 concentrations
simulated by the GEOS-Chem and CAMx base case models.
4th-highest MDA8 O3
Region Observation Model concentration (ppbv)a
California (5 sites) CASTNET0
GEOS-Chem (paired)d
GEOS-Chem (unpaired)6
CAMx (paired)d
CAMx (unpaired)6
West (1 4 sites) CASTNET0
GEOS-Chem (paired)d
GEOS-Chem (unpaired)6
CAMx (paired)'
CAMx (unpaired)6
North Central (6 sites) CASTNET0
GEOS-Chem (paired)d
GEOS-Chem (unpaired)6
CAMx (paired)d
CAMx (unpaired)6
Northeast (5 sites) CASTNET0
GEOS-Chem (paired)d
GEOS-Chem (unpaired)6
CAMx (paired)'
CAMx (unpaired)6
Southeast (9 sites) CASTNET0
GEOS-Chem (paired)'
GEOS-Chem (unpaired)6
CAMx (paired)'
CAMx (unpaired)6
90 ±13°
71 ±15
85 ± 19
71 ±9
85 ±13
70 ±4°
62 ±8
68 ±7
63 ±8
71 ±7
71 ±5°
58 ±10
69 ± 10
63 ±7
73 ±8
71 ±4°
61 ±6
68 ±5
72 ±7
75 ±3
76 ±8°
61 ±6
71 ±5
71 ± 11
79 ±9
Number of Days
within 5 ppb
0
0
4
6
1
1
0
3
2
5
"Annual 4th-highest (99th-percentile) MDA8 O3 concentration regional means (ppb) ± standard deviation;
bNumberof days the model predicted MDA8 O3 concentrations were within 5 ppb of the observed 4th-highest concentrations;
°Observed concentrations at CASTNET sites;
'Modeled concentrations at CASTNET sites on days when the 4th-highest MDA8 O3 concentration was observed (paired by date);
6Model predicted annual 4th-highest MDA8 O3 concentration at CASTNET sites (unpaired by date).
Source: Data from Zhang etal. (2011) for GEOS-Chem and Emery etal. (2012) for CAMx.
3-61
-------
Comparison of GEOS-Chem results for natural and NA background indicate that
methane is also a major contributor to NA background O3, accounting for slightly
less than half of the increase in background since the pre-industrial era and whose
relative contribution is projected to grow in the future. U.S. background
concentrations are on average 2.6 ppb higher than NA background concentrations
during spring and 2.7 ppb during summer across the United States. Highest values
for U.S. background (in the U.S.) are found over the Northern Tier of New York
State (19.1 ppb higher than local NA background concentrations) in summer. High
values are also found in other areas bordering Canada and Mexico.
Analyses of results from GEOS-Chem and CAMx presented here and shown in
Table 3-1 and Table 3-2 are in accord with results from Kasibhatla and Chameides
(2000) who found that the accuracy of simulations improved as the averaging time of
both the simulation and the observations increased (see Section 3.3). Note that any
CTM—not just the ones considered here—will have difficulty in predicting day
specific quantities. When analyzing results over long time periods (e.g., months),
special care should be taken to examine temporal trends in bias because this will
improve understanding of the modeling results.
Overall, these results suggest that GEOS-Chem is capable of simulating seasonal or
monthly mean MDA8 O3 to within a few parts per billion on a regional basis
throughout the U.S., except in California. These results suggest that CAMx is
capable of simulating seasonal or monthly mean MDA8 O3 to within a few ppb,
though, CAMx also shows relatively large disagreements in California and, in
addition, shows relatively large positive bias in seasonal mean MDA8 O3 in the
eastern United States. However, differences between the models in the East are likely
to narrow with updates to chemistry. Neither model is capable of simulating
4th-highest MDA8 O3 to within suitable bounds on a day-specific basis at all sites, or
even most sites. However, agreement between simulated versus observed 4th-highest
MDA8 O3 is improved for either model when the models and the measurements are
sampled on different days.
Note that the calculations of background concentrations presented in this section
were formulated to answer the question, "what would O3 concentrations be if there
were no anthropogenic sources." This is different from asking, "how much of the O3
measured or simulated in a given area is due to background contributions." Because
of potentially strong non-linearities (i.e., the fate, or lifetime, of the background O3
transported into the urban area will depend on the concentration of the background
O3 in addition to interactions of background O3 with the local chemical regime) in
many urban areas, these estimates by themselves should not be used to answer the
second question posed above. The extent of these non-linearities will generally
depend on location and time, the strength of concentrated sources and the nature of
the chemical regime.
3-62
-------
3.5 Monitoring
3.5.1 Routine Monitoring Techniques
The federal reference method (FRM) for O3 measurement is called the
Chemiluminescence Method (CLM) and is based on the detection of
chemiluminescence resulting from the reaction of O3 with ethylene gas. The UV
absorption photometric analyzers were approved as federal equivalent methods
(FEMs) in 1977 and gained rapid acceptance for NAAQS compliance purposes due
to ease of operation, relatively low cost, and reliability. The UV absorption method is
based on the principle that O3 molecules absorb UV radiation at a wavelength of
254 nm from a mercury lamp. The concentration of O3 is computed from Beer's law
using the radiation absorbed across a fixed path length, the absorption coefficient,
and the measured pressure and temperature in the detection cell. UV absorption
photometry is the predominant method for assessing compliance with the NAAQS
for O3. Almost all of the state and local air monitoring stations (SLAMS) that
reported data to EPA AQS from 2005 to 2009 used UV absorption photometer
FEMs. No CLM monitors, approved as FRMs or FEMs, reported O3 data to AQS
from 2005 to 2009 and only one monitor reported data using a long-path or open path
Differential Optical Absorption Spectrometer (DOAS) FEM during this period.
The rationale, history, and calibration of O3 measurements were summarized in the
1996 and 2006 O3 AQCDs (U.S. EPA. 2006b. 1996a) and focused on the state of
ambient O3 measurements at that time as well as evaluation of interferences and new
developments. This discussion will continue with the current state of O3
measurements, interferences, and new developments for the period 2005 to 2010.
UV O3 monitors use mercury lamps as the source of UV radiation and employ an O3
scrubber (typically manganese dioxide) to generate an O3-free air flow to serve as a
reference channel for O3 measurements. There are known interferences with UV O3
monitors. The 2006 O3 AQCD (U.S. EPA. 2006b) reported on the investigation of
the effects of water vapor, aromatic compounds, ambient particles, mercury vapor
and alternative materials in the instrument's O3 scrubber. The overall conclusions
from the review of the scientific literature covered in the 2006 O3 AQCD (U.S. EPA,
2006b) are briefly summarized below.
Kleindienst et al. (1993) found water vapor to have no measurable impact and
aromatic compounds to have a minor impact (as much as 3% higher than the FRM
extrapolated to ambient conditions) on UV absorption measurements. UV O3
monitor response evaluated by chamber testing using cigarette smoke, reported an
elimination of the O3 monitor response to the smoke when a particle filter was used
that filtered out particles less than 0.2 j^m in diameter (Arshinov et al., 2002). One
study (Leston et al., 2005) in Mexico City compared a UV O3 FEM to a CLM FRM.
The UV FEM reported consistently higher O3 than the CLM FRM. They suggested
that O3 measured in ambient air could be too high by 20 to 40 ppb under specific
conditions due to positive interference by a number of organic compounds, mainly
3-63
-------
those produced during the oxidation of aromatic hydrocarbons and some primary
compounds such as styrene and naphthalene. However, the concentrations of these
compounds were many times higher in both of these environments than are typically
found at ambient air monitoring sites in the United States Although Hg is also
potentially a strong interfering agent, because the Hg resonance line is used in this
technique, its concentration would also have to be many times higher than is
typically found in ambient air, e.g., as might be found in power plant plumes. Thus, it
seems unlikely that such interferences would amount to more than one or two ppb
(within the design specifications of the FEM), except under conditions conducive to
producing high concentrations of the substances they identified as causing
interference. Leston et al. (2005) also presented smog chamber data which
demonstrated that heated metal and heated silver wool scrubbers perform better in
the presence of aromatic hydrocarbon irradiations than manganese dioxide scrubbers
when compared to the FRM. They also suggested the use of humidified calibration
gas and alternative scrubber materials to improve UV O3 measurements. Some O3
monitor manufacturers now offer heated silver wool scrubbers as an alternative to
manganese dioxide. Another possible solution to the O3 scrubber problem may be the
use of a gas phase scrubber such as NO. A commercial version of this has recently
been introduced by 2B Technologies as an option on their model 202 FEM; however,
it has not been field tested or approved for use as an FEM.
Review of the recent literature is summarized below. Study of UV monitors by
Williams et al. (2006) concluded that well maintained monitors showed no
substantial interferences when operated in locations with high concentrations of
potentially interfering VOCs including Nashville, Houston, and the Gulf of Maine.
Monitors were tested in urban and suburban environments, as well as on board a ship
in both polluted and clean marine air. Comparisons of UV measurements to a non-
FRM/FEM NO based CLM demonstrated agreement to within 1%. At the Houston
location, they did observe a brief period on one day for about 30 minutes where the
UV measurements exceeded the CLM by about 8 ppb (max). This was attributed to
probable instrument malfunction.
Wilson and Birks (2006) investigated water vapor interference in O3 measurements
by four different UV monitors. In extreme cases where a rapid step change in relative
humidity between 0 and 90% was presented, large transitory responses (tens to
hundreds of ppb) were found for all monitors tested. Rapid changes in relative
humidity such as this would not be expected during typical ambient O3
measurements and could only be expected during measurement of vertical profiles
from balloon or aircraft. The magnitude of the interference and the direction (positive
or negative) was dependent on the manufacturer and model. Wilson and Birks (2006)
also hypothesized that water vapor interference is caused by physical interactions of
water vapor on the detection cell. The O3 scrubber was also thought to act as a
reservoir for water vapor and either added or removed water vapor from the air
stream, subsequently affecting the detector signal and producing either a positive or
negative response. They demonstrated that the use of a Nafion permeation membrane
just before the O3 detection cell to remove water vapor eliminated this interference.
3-64
-------
Dunlea et al. (2006) evaluated multiple UV O3 monitors with two different O3
scrubber types (manganese dioxide and heated metal wool) in Mexico City. Large
spikes in O3 concentrations were observed while measuring diesel exhaust where
large increases in particle number density were observed. The interference due to
small particles passing through the Teflon filter and scattering/absorbing light in the
detection cell were estimated to cause at most a 3% increase in measurements in
typical ambient air environments. This estimate pertains to measurements in the
immediate vicinity of fresh diesel emissions and most monitor siting guidelines
would not place the monitor close to such sources, so actual interferences are
expected to be much less than 3%. Dunlea et al. (2006) also observed no evidence for
either a positive or negative interference or dependence due to variations in aromatics
during their field study.
Li et al. (2006c) verified early reports of gas phase mercury interference with the UV
O3 measurement. They found that 300 ng/m3 of mercury produced an instrument
response of about 35 ppb O3. Background concentrations of mercury are around 1-
2 ng/m3 and expected to produce an O3 response that would be <1 ppb.
Spicer et al. (2010) examined potential UV O3 monitor interferences by water vapor,
mercury, aromatic compounds, and reaction products from smog chamber
simulations. Laboratory tests showed little effect of changing humidity on
conventional FEM UV O3 monitors with manganese dioxide or heated metal wool
scrubbers in the absence of other interferences. Mercury vapor testing produced an
O3 response by the UV monitors that was <1 ppb O3 per 1 ppt (about 8 ng/m3)
mercury vapor. Interference by aromatic compounds at low (3% RH) and high (80%
RH) humidity showed some positive responses that varied by UV monitor and
ranged from 0 to 2.2 ppb apparent O3 response, per ppb of aromatic compound
tested. The authors acknowledged that the aromatic compounds most likely to
interfere are rarely measured in the atmosphere and therefore, make it difficult to
assess the impact of these compounds during ambient air monitoring. Comparison of
UV and CLM responses to photochemical reaction products in smog chamber
simulations at 74% to 85% RH showed varied responses under low
(0.125 ppmv/0.06 ppmv) to high (0.50 ppmv/0.19 ppmv) hydrocarbon/NOx
conditions. The conventional UV monitors were as much as 2 ppb higher than the
CLM under low hydrocarbon/NOx conditions and 6 ppb higher under the high
hydrocarbon/NOx conditions. Two FEM UV monitors were also co-located at six
sites in Houston from May to October, 2007 with one UV monitor equipped with
Nafion permeation membrane. The average difference between 8-h daily max O3
concentrations using the UV and the UV with Nafion permeation membrane ranged
from -4.0 to 4.1 ppb.
3-65
-------
3.5.2 Precision and Bias
In order to provide decision makers with an assessment of data quality, EPA's
Quality Assurance (QA) group derives estimates of both precision and bias for O3
and the other gaseous criteria pollutants from the biweekly single point quality
control (QC) checks using calibration gas, performed at each site by the monitoring
agency. The single-point QC checks are typically performed at concentrations around
90 ppb. Annual summary reports of precision and bias can be obtained for each
monitoring site at http://www.epa.gov/ttn/amtic/qareport.html. The assessment of
precision and bias are based on the percent-difference values, calculated from single-
point QC checks. The percent difference is based on the difference between the
pollutant concentration indicated by monitoring equipment and the known (actual)
concentration of the standard used during the QC check. The monitor precision is
estimated from the 90% upper confidence limit of the coefficient of variation (CV) of
relative percent difference (RPD) values. The bias is estimated from the 95% upper
confidence limit on the mean of the absolute values of percent differences. The data
quality goal for O3 precision and bias at the 90 and 95% upper confidence limits is
7% (40 CFR Part 58, Appendix A). Table 3-3 presents a summary of the number of
monitors that meet the precision and bias goal of 7% for 2005 to 2009. Greater than
96% of O3 monitors met the precision and bias goal between 2005 and 2009.
Another way to look at the precision (CV) and bias (percent difference) information
using the single-point QC check data from the monitoring network is to present box
plots of the monitors' individual precision and percent-difference data; Figure 3-17
and Figure 3-18 include this information for O3 monitors operating from 2005 to
2009.
Table 3-3 Summary of O3 monitors meeting 40 CFR Part 58, Appendix A
Precision and Bias Goals.
Monitors with Monitors with
Year Number of Monitors Acceptable Precision (%) Acceptable Bias (%)
2005 879 96.5 96.7
2006 881 98.1 97.6
2007 935 98.1 98.1
2008 955 97.1 96.7
2009 958 97.4 97.5
3-66
-------
90" percent! I e
75" percentile
Mean
Median
25" percentile
10" percentile
I
2005
N=1151
2006
N=1159
2007
N=1166
2008
N=1178
2009
N=11S8
Figure 3-17 Box plots of precision data by year (2005-2009) for all Os monitors
reporting single-point QC check data to AQS.
10-
-
6
5-
4
3
2
1 •
0 -
-1 •
-2-
-3-
-4
-5
•t
•1
-C-:
•9
ir -
on" percentile
75IH percentile
Mean
Median
25IH percentile
10" percentile
Legend
I
I
200S
N=52724
2006
N=51814
2007
N=53262
2008
N=57315
2009
N=67305
Figure 3-18 Box plots of percent-difference data by year (2005-2009) for all
Os monitors reporting single-point QC check data to AQS.
3-67
-------
3.5.2.1 Precision from Co-located UV Os Monitors in Missouri
The Missouri Department of Natural Resources (MODNR) maintains a network of
co-located UV O3 analyzers. The MODNR provided co-located data from four
monitors: two co-located at the same monitoring site in Kansas City (AQS ID
290370003) and two co-located at the same monitoring site in St. Louis (AQS ID
291831002). Hourly observations for the co-located measurements at these two sites
between April and October, 2006-2009 were used to evaluate precision from
co-located UV monitors. These data were then compared with the precision obtained
by the biweekly single point QC checks for all sites reporting single-point QC check
data to AQS between 2005 and 2009; the method normally used for assessing
precision. Box plots of the RPD between the primary and co-located hourly O3
measurements in Missouri are shown in Figure 3-19 and box plots of the RPD
between the actual and indicated QC check for all U.S. sites are shown in
Figure 3-20. As mentioned above, the average concentration of the single-point QC
check is 90 ppb, whereas the average ambient O3 concentration measured at the two
sites in Missouri was 34 ppb. The mean RPD for the co-located monitors in Missouri
and the single-point QC check data from all sites were less than 1 percent.
_ •
-3-
2006
N=10017
Legend
90" percentile
75" | len enhle
Mean
Median
25"'percentile
10" pelccnillc
2007
N=10133
2008
N=9884
2009
N=10211
Figure 3-19 Box plots of RPD data by year for the co-located O3 monitors at
two sites in Missouri from 2006-2009.
3-68
-------
Relative Percent Difference
U> hj -1* O -i Ixj (jJ
III
III
Leg
90" percentile -
75 u' percentile -
Mean
Median ~
25" percentile -
10lh percentile _
I
I
end
r
h
r
I
-|-
2
2005
N=52724
2006
N=51814
2007
N=53262
2008
N=S7315
2009
N=67305
Figure 3-20 Box plots of RPD data by year for all of the United States.
Ozone sites reporting single-point QC check data to AQS from
2005-2009.
3.5.3 Performance Specifications
The performance specifications for evaluating and approving new FEMs in
accordance with 40 CFR Part 53 are provided in Table 3-4. These specifications were
developed and originally published in the Federal Register in 1975. Modern,
commercially-available instruments can now perform much better than the
requirements specified below. For example, the lower detectable limit (LDL)
performance specification is 10 ppb and the typical vendor-stated performance for
the LDL is now less than 0.60 ppb. The amount of allowable interference equivalent
for total interference substances is 60 ppb, and the current NAAQS for O3 is 75 ppb,
with an averaging time of 8 hours. Improvements in new measurement technology
have occurred since these performance specifications were originally developed.
These specifications should be revised to more accurately reflect the necessary
performance requirements for O3 monitors used to support the current NAAQS.
3-69
-------
Table 3-4 Performance specifications for O$ based in 40 CFR Part 53.
Parameter Specification
Range 0 - 0.5 ppm (500 ppb)
Noise 0.005 ppm (5 ppb)
LDL - defined as two times the noise 0.01 ppm (10 ppb)
Interference equivalent
Each interfering substance ± 0.02 ppm (20 ppb)
Total interfering substances 0.06 ppm (60 ppb)
Zero drift
12 h ± 0.02 ppm (20 ppb)
24 h ± 0.02 ppm (20 ppb)
Span Drift, 24 h
20% of upper range limit
80% of upper range limit
Lag time
Rise time
Fall time
± 20.0%
± 5.0%
20 min
15 min
15 min
Precision
20% of upper range limit
80% of upper range limit
0.01 ppm (10 ppb)
0.01 ppm (10 ppb)
3.5.4 Monitor Calibration
The calibration of O3 monitors was summarized in detail in the 1996 O3 AQCD
(U.S. EPA, 1996a). The calibration of O3 monitors is done using an O3 generator and
UV photometers. UV photometry is the prescribed procedure for the calibration of
reference methods to measure O3 in the atmosphere. Because O3 is unstable and
cannot be stored, the O3 calibration procedure specifically allows the use of transfer
standards for calibrating ambient O3 monitors. A transfer standard is calibrated
against a standard of high authority and traceability and then moved to another
location for calibration of O3 monitors. The EPA and the National Institute of
Standards and Technology (NIST) have established a network of standard reference
photometers (SRPs) that are used to verify transfer standards. The International
Bureau of Weights and Measures (BIPM) maintain one NIST SRP (SRP27) as the
World's O3 reference standard. NIST maintains two SRPs (SRPO and SRP2) that are
used for comparability to ten other SRPs maintained by the EPA's Regional QA
staff.
SRPs have been compared to other reference standards. Tanimoto et al. (2006)
compared NIST SRPS5, owned by the National Institute for Environmental Studies
3-70
-------
in Japan, to gas phase titration (GPT). The SRP was found to be 2% lower than GPT.
GPT is no longer used as a primary or transfer standard in the United States. Viallon
et al. (2006) compared SRP27 built at BIPM to four other NIST SRPs maintained by
BIPM (SRP28, SRP31, SRP32, and SRP33). A minimum bias of+0.5% was found
for all SRP measurement results, due to use of the direct cell length measurement for
the optical path length; this bias was accounted for by applying the appropriate
correction factor. Study of the bias-corrected SRPs showed systematic biases and
measurement uncertainties for the BIPM SRPs. A bias of -0.4% in the instrument O3
mole fraction measurement was identified and attributed to non-uniformity of the gas
temperature in the instrument gas cells, which was compensated by a bias of+0.5%
due to an under-evaluation of the UV light path length in the gas cells. The relative
uncertainty of the O3 absorption cross section was 2.1% at 253.65 nm and this was
proposed as an internationally accepted consensus value until sufficient experimental
data is available to assign a new value.
In November, 2010, the EPA revised the Technical Assistance Document for
Transfer Standards for Calibration of Air Monitoring Analyzers for Ozone (U.S.
EPA. 201 Of) that was first finalized in 1979 (U.S. EPA. 1979b). The revision
removed methods no longer in use and updated definitions and procedures where
appropriate. In the revised document, the discussion of transfer standards for O3
applies to the family of standards that are used beyond SRPs or Level 1 standards.
To reduce confusion, EPA reduced the number of common terms that were used in
the past such as: primary standard, local primary standard, transfer standard, and
working standard. Beyond the SRPs, all other standards are considered transfer
standards.
3.5.5 Other Monitoring Techniques
3.5.5.1 Portable UV O3 Monitors
Small, lightweight, and portable UV O3 monitors with low power consumption are
commercially available. These monitors are based on the same principle of UV
absorption by O3 at 254 nm. Monitors of this type are typically used for vertical
profiling using balloons, kites, or light aircraft where space and weight are limited.
They have also been used for monitoring at remote locations such as National Parks.
Burley and Ray (2007) compared portable O3 monitor measurements to those from a
conventional UV monitor in Yosemite National Park. Calibrations of the portable O3
monitors against a transfer standard resulted in an overall precision of ± 4 ppb and
accuracy of ± 6%. Field measurement comparisons between the portable and
conventional monitor at Turtleback Dome showed the portable monitor to be 3.4 ppb
lower on average, with daytime deviation typically on the order of 0-3 ppb.
Agreement between the portable and conventional monitor during daylight hours
(9:00 a.m. to 5:00 p.m. PST) resulted in an R2 of 0.95, slope of 0.95, and intercept of
0.36 ppb. Substantial deviations were observed in the predawn hours where the
3-71
-------
portable monitor was consistently low. These deviations were attributed to the
difference in sampling inlet location. The portable monitor was located at 1.3 meters
above ground and the conventional monitor was located at 10 meters above ground.
Agreement between the portable and conventional monitors for all hours sampled
resulted in an R2 of 0.88, slope of 1.06, and intercept of-6.8 ppb. (Greenberg et al..
2009) also compared a portable UV O3 monitor to a conventional UV monitor in
Mexico City and obtained good agreement for a 14 day period with an R2 of 0.97,
slope of 0.97, and intercept of 6 ppb. One portable O3 monitor was recently approved
as an FEM (EQOA-0410-190) on April 27, 2010 (75 FR 22126).
3.5.5.2 NO-based Chemiluminescence Monitors
One commercially available NO-based chemiluminescence monitor has been
approved as an FEM (EQOA-0611-199) on October 7, 2011 (75 FR 62402). It may
also be designated as a second or replacement FRM since the ethene based FRMs are
no longer manufactured. Although this is a relatively new monitor, other NO-based
CLM instruments have been custom built for various field studies since the early
1970s. A commercial version that measured both O3 and NOX was offered in the
early 1970s but failed to gain commercial acceptance. Initial testing with SO2, NO2,
C12, C2H2, C2H4 and C3H6 (Stedman et al., 1972) failed to identify any
interferences. In the intervening years, custom built versions have not been found to
have any interference; however, they do experience a slight decrease in response with
increasing relative humidity (due to quenching of the excited species by the water
molecules). The new NO-based CLM solves this problem with the use of a Nafion
membrane dryer. A custom built NO-based CLM similar to the FEM was used by
Williams et al. (2006) in Houston, TX; Nashville, TN; and aboard ship along the
New England coast. It was found to be in good agreement with a standard UV based
FEM and with a custom built DOAS.
3.5.5.3 Passive Air Sampling Devices and Sensors
A passive O3 sampling device depends on the diffusion of O3 in air to a collecting or
indicating medium. In general, passive samplers are not adequate for compliance
monitoring because of the limitations in averaging time (typically one week or
more), particularly for O3. However, these devices are valuable for personal human
exposure estimates and for obtaining long-term data in rural areas where
conventional UV monitors are not practical or feasible to deploy. The 1996 O3
AQCD (U.S. EPA, 1996a) provided a detailed discussion of passive samplers, along
with the limitations and uncertainties of the samplers evaluated and published in the
literature from 1989 to 1995. The 2006 O3 AQCD (U.S. EPA. 2006b) provided a
brief update on available passive samplers developed for use in direct measurements
of personal exposure published through 2004. The 2006 O3 AQCD (U.S. EPA.
2006b) also noted the sensitivity of these samplers to wind velocity, badge
3-72
-------
placement, and interference by other copollutants that may result in measurement
error.
Subsequent evaluations of passive diffusion samplers in Europe showed good
correlation when compared to conventional UV O3 monitors, but a tendency for the
diffusion samplers to overestimate the O3 concentration (Gottardini et al.. 2010:
Vardoulakis et al., 2009; Buzica et al., 2008). The bias of O3 diffusion tubes were
also found to vary with concentration, season, and exposure duration (Vardoulakis et
al., 2009). Development of simple, inexpensive, passive O3 measurement devices
that rely on O3 detection papers and a variety of sensors with increased time
resolution (sampling for hours instead of weeks) and improved sensitivity have been
reported (Maruo et al., 2010; Ebeling et al., 2009; Miwa et al., 2009; Ohira et al.,
2009; Maruo, 2007; Utembe et al., 2006). Limitations for some of these sensors and
detection papers include air flow dependence and relative humidity interference.
3.5.5.4 Differential Optical Absorption Spectrometry
Optical remote sensing methods can provide direct, sensitive, and specific
measurements of O3 over a broad area or open path in contrast with conventional
single-point UV monitors. The 1996 O3 AQCD (U.S. EPA. 1996a) provided a brief
discussion of DOAS for O3 measurements and cited references to document the
sensitivity (1.5 ppb for a 1-minute averaging time), correlation (r = 0.89), and
agreement (on the order of 10%) with UV O3 monitors (Stevens et al.. 1993).
The 2006 O3 AQCD (U.S. EPA. 2006b) provided an update on DOAS where a
positive interference due to an unidentified absorber was noted (Reisinger. 2000).
More recent study of the accuracy of UV absorbance monitors by Williams et al.
(2006) compared UV and DOAS measurements at two urban locations. In order to
compare the open path measurements and UV, the data sets were averaged to 30-
minute periods and only data when the boundary layer was expected to be well
mixed (between 10:00 a.m. and 6:00 p.m. CST) were evaluated. The comparisons
showed variations of no more than ± 7% (based on the slope of the linear least
squares regression over a concentration range from about 20 to 200 ppb) and good
correlation (R2 = 0.96 and 0.98). Lee et al. (2008b~) evaluated DOAS and UV O3
measurements in Korea and found the average DOAS concentration to be 8.6%
lower than the UV point measurements with a good correlation (R2 = 0.94).
DOAS has also been used for the measurement of HNO2 (or HONO). DOAS was
compared to chemical point-measurement methods for HONO. Acker et al. (2006)
obtained good results when comparing wet chemical and DOAS during well mixed
atmospheric conditions (wet chemical = 0.009 + 0.92 x DOAS; r = 0.7). Kleffmann
and Wiesen (2008) noted that interferences with the HONO wet chemical methods
can affect results from inter-comparison studies if not addressed. In an earlier study,
Kleffmann et al. (2006) demonstrated that when the interferences were addressed,
excellent agreement with DOAS can be obtained. Stutz et al. (2010) found good
agreement (15% or better) between DOAS and a wet chemical method (Mist
3-73
-------
Chamber/Ton Chromatography) in Houston, TX except generally during mid-day
when the chemical method showed a positive bias that may have been related to
concentrations of O3. DOAS remains attractive due to its sensitivity, speed of
response, and ability to simultaneously measure multiple pollutants; however, further
inter-comparisons and interference testing are recommended.
3.5.5.5 Satellite Remote Sensing
Satellite observations for O3 are growing as a resource for many purposes, including
model evaluation, assessing emissions reductions, pollutant transport, and air quality
management. Satellite remote sensing instruments do not directly measure the
composition of the atmosphere. Satellite retrievals are conducted using the solar
backscatter or thermal infrared emission spectra and a variety of algorithms. Most
satellite measurement systems have been developed for stratospheric measurement of
the total O3 column. Mathematical techniques have been developed and must be
applied to derive information from these systems about tropospheric O3 (Tarasick
and Slater, 2008; Ziemke et al., 2006). Direct retrieval of global tropospheric O3
distributions from solar backscattered UV spectra have been reported from OMI and
the Global Ozone Monitoring Experiment (GOME) (Liu et al., 2006). Another
satellite measurement system, Tropospheric Emission Spectrometer (TES), produces
global-scale vertical concentration profiles of tropospheric O3 from measurements of
thermal infrared emissions. TES has been designed specifically to focus on mapping
the global distribution of tropospheric O3 extending from the surface to about 10-
15 km altitude (Beer, 2006). Satellite measurements of tropospheric O3 generally
require monthly averaging to reduce noise, and thus are of little use for observing
synoptic-scale variability.
In order to improve the understanding of the quality and reliability of the data,
satellite-based observations of total column and tropospheric O3 have been validated
in several studies using a variety of techniques, such as aircraft observations,
ozonesondes, CTMs, and ground-based spectroradiometers. Anton et al. (2009)
compared satellite data from two different algorithms (OMI-DOAS and OMI-
TOMS) with total column O3 data from ground-based spectroradiometers at five
locations. The satellite total column O3 data underestimated ground-based
measurements by less than 3%. Richards et al. (2008) compared TES tropospheric
O3 profiles using airborne differential absorption lidar (DIAL) and found TES to
have a 7 ppbv positive bias relative to DIAL throughout the troposphere. Nassar et al.
(2008) compared TES O3 profiles and ozonesonde coincidences and found a positive
bias of 3-10 ppbv for TES. Worden et al. (2007a) also compared TES with
ozonesondes and found TES O3 profiles to be biased high in the upper troposphere
(average bias of 16.8 ppbv for mid-latitudes and 9.8 ppbv for the tropics) and biased
low in the lower troposphere (average bias of -2.6 ppbv for mid-latitudes and -
7.4 ppbv for the tropics). Comparisons of TES and OMI with ozonesondes by Zhang
et al. (201 Ob) showed a mean positive bias if 5.3 ppbv (10%) for TES and 2.8 ppbv
(5%) for OMI at 500 hPa. In addition, Zhang et al. (201 Ob) used a CTM (GEOS-
3-74
-------
Chem) to determine global differences between TES and OMI. They found
differences between TES and OMI were generally ±10 ppbv except at northern mid-
latitudes in summer and over tropical continents. Satellite observations have also
been combined (e.g., OMI and TES) to improve estimates of tropospheric O3
(Worden et al.. 2007b).
Satellite measurements are also available for O3 precursors such as CO, NO2, and
HCHO. These measurements are useful for constraining model estimates of O3
precursor emissions and long-range transport of pollution (Section 3.4). Zhang et al.
(2008) used satellite measurements of CO and NO2 along with O3 to constrain
estimates of background O3 precursors in Asia and O3 produced during long range
transport.
3.5.6 Ambient O3 Network Design
3.5.6.1 Monitor Siting Requirements
To monitor compliance with the NAAQS, state and local monitoring agencies
operate O3 monitoring sites at various locations depending on the area size
(population and geographic characteristics1) and typical peak concentrations
(expressed in percentages below, or near the O3 NAAQS). SLAMS make up the
ambient air quality monitoring sites that are primarily needed for NAAQS
comparisons, but may also serve some other basic monitoring objectives that include:
providing air pollution data to the general public in a timely manner; emissions
strategy development; and support for air pollution research. SLAMS include
National Core (NCore), Photochemical Assessment Monitoring Stations (PAMS),
and all other State or locally-operated stations except for the monitors designated as
special purpose monitors (SPMs).
The SLAMS minimum monitoring requirements to meet the O3 design criteria are
specified in 40 CFR Part 58, Appendix D. Although NCore and PAMS are a subset
of SLAMS, the monitoring requirements for those networks are separate and
discussed below. The minimum number of O3 monitors required in a Metropolitan
Statistical Area (MSA) ranges from zero for areas with a population of at least
50,000 and under 350,000 with no recent history of an O3 design value2 greater than
85 percent of the NAAQS, to four for areas with a population greater than 10 million
and an O3 design value greater than 85 percent of the NAAQS. Within an O3
network, at least one site for each MSA, or Combined Statistical Area (CSA) if
multiple MSAs are involved, must be designed to record the maximum concentration
for that particular metropolitan area. More than one maximum concentration site may
1 Geographic characteristics such as complexity of terrain, topography, land use, etc.
2 A design value is a statistic that describes the air quality status of a given area relative to the level of the NAAQS. Design values
are typically used to classify nonattainment areas, assess progress toward meeting the NAAQS, and develop control strategies.
See http://epa.aov/airtrends/values.html (U.S. EPA. 201 Oa) for guidance on how these values are defined.
3-75
-------
be necessary in some areas. The spatial scales for O3 sites are neighborhood, urban
and regional.
• Neighborhood scale: represents concentrations within some extended area of
the city that has relatively uniform land use with dimensions in the 0.5-4.0 km
range. The neighborhood and urban scales listed below have the potential to
overlap in applications that concern secondary or homogeneously distributed
primary air pollutants.
• Urban scale: represents concentrations within an area of city-like dimensions,
on the order of 4-50 km. Within a city, the geographic placement of sources
may result in there being no single site that can be said to represent air quality
on an urban scale.
• Regional scale: usually defines a rural area of reasonably homogeneous
geography without large sources, and extends from tens to hundreds of
kilometers.
Since O3 concentrations decrease appreciably in the colder parts of the year in many
areas, O3 is required to be monitored at SLAMS monitoring sites only during the
"ozone season." Table D-3 of 40 CFR Part 58, Appendix D lists the beginning and
ending month of the ozone season for each U.S. state or territory. Most operate O3
monitors only during the ozone season. Those that operate some or all of their O3
monitors on a year-round basis include Arizona, California, Hawaii, Louisiana,
Nevada, New Mexico, Puerto Rico, Texas, American Samoa, Guam and the Virgin
Islands.
The total number of SLAMS O3 sites needed to support the basic monitoring
obj ectives includes more sites than the minimum numbers required in 40 CFR Part
58, Appendix D. In 2010, there were 1250 O3 monitoring sites reporting values to the
EPA AQS database (Figure 3-21). Monitoring site information for EPA's air quality
monitoring networks is available in spreadsheet format (CSV) and keyhole markup
language format (KML or KMZ) that is compatible with Google Earth™ and other
software applications on the AirExplorer website (U.S. EPA. 201 le). States may
operate O3 monitors in non-urban or rural areas to meet other objectives
(e.g., support for research studies of atmospheric chemistry or ecosystem impacts).
These monitors are often identified as SPMs and can be operated up to 24 months
without being considered in NAAQS compliance determinations. The current
monitor and probe siting requirements have an urban focus and do not address the
siting for SPMs or monitors in non-urban, rural areas to support ecosystem impacts
and the secondary standards.
3-76
-------
o Urban NCore
o PA MS
• Other Sites Reporting to AQS
0 250 500 1000 Miles D 55110 220 Mies 0 250 600 1 COO Miles
Puerto Rico &
Virgin Islands
Alaska
D 25 5D 100 Miles
Figure 3-21 U.S. O3 sites reporting data to AQS in 2010.
NCore is a new multipollutant monitoring network implemented to meet multiple
monitoring objectives. Those objectives include: timely reporting of data to the
public through AirNow (U.S. EPA, 201 la); support for the development of emission
reduction strategies; tracking long-term trends of criteria pollutants and precursors;
support to ongoing reviews of the NAAQS and NAAQS compliance; model
evaluation; support for scientific research studies; and support for ecosystem
assessments. Each state is required to operate at least one NCore site. The NCore
monitoring network began January 1, 2011 at about 80 stations (about 60 urban and
20 rural sites). NCore has leveraged the use of sites in existing networks; for
example, some IMPROVE sites also serve as rural NCore sites. In addition to O3,
other components including CO, NOX, NOY, SO2, and basic meteorology are also
measured at NCore sites. The spatial scale for urban NCore stations is urban or
neighborhood; however, a middle-scale1 site may be acceptable in cases where the
site can represent many such locations throughout a metropolitan area. Rural NCore
sites are located at a regional or larger scale, away from any large local emission
sources so that they represent ambient concentrations over an extensive area. Ozone
monitors at NCore sites are operated year round.
1 Middle scale defines an area up to several city blocks in size with dimensions ranging from about 100 to 500 meters.
3-77
-------
PAMS provides more comprehensive data on O3 in areas classified as serious,
severe, or extreme nonattainment for O3. In addition to O3, PAMS provides data for
NOX, NOY, VOCs, carbonyls, and meteorology. The PAMS network design criteria
are based on locations relative to O3 precursor source areas and predominant wind
directions associated with high O3 concentrations. The overall network design is
location specific and geared toward enabling characterization of precursor emission
sources in the area, O3 transport, and photochemical processes related to O3
nonattainment. Minimum monitoring for O3 and its precursors is required annually
during the months of June, July, and August when peak O3 concentrations are
expected. In 2006, the EPA reduced the minimum PAMS monitoring requirements
(71 FR 61236). There were a total of 92 PAMS sites reporting values to the AQS
database in 2010.
CASTNET is a regional monitoring network established to assess trends in acidic
deposition due to emission reduction regulations. CASTNET also provides
concentration measurements of air pollutants involved in acidic deposition, such as
sulfate and nitrate, in addition to the measurement of O3. CASTNET O3 monitors
operate year round and are primarily located in rural areas. In 2010, there were 80
CASTNET sites located in, or near, rural areas. As part of CASTNET, the National
Park Service (NFS) operates 23 sites located in national parks and other Class-I
areas. Ozone data collected at the 23 NFS sites is compliant with the SLAMS QA
requirements in 40 CFR Part 58, Appendix A. Ozone measurements at the remaining
CASTNET sites were not collected with the QA requirements for SLAMS outlined
in 40 CFR Part 58, Appendix A, and therefore, these O3 data cannot be used for
NAAQS compliance purposes. The SLAMS QA requirements and procedures are
currently being implemented at the remaining sites.
The NFS also operates a Portable Ozone Monitoring Systems (POMS) network.
The POMS couples the small, low-power O3 monitor with a data logger,
meteorological measurements, and solar power in a self contained system for
monitoring in remote locations. Typical uses for the POMS data include research
projects, survey monitoring, and assessments of spatial O3 distribution. The portable
O3 monitor in use by the NFS was recently designated as an equivalent method for
O3 (75 FR 22126). Seventeen NFS POMS monitors were operating in 2010 (NFS.
2011). A map of the rural NCore sites, along with the CASTNET, and the NFS
POMS sites are shown in Figure 3-22. As can be seen from Figure 3-21 and
Figure 3-22, vast rural areas of the country still exist without any monitor coverage.
Monitoring opportunities exist in these areas where relatively few and easily
characterized precursor sources dominate and could be used to improve
understanding of O3 formation.
3-78
-------
Alaska
D 25D 5DD
lii ill
1JIE Mfct D K11D 22] lilts
Rural HCore
HPS POMS
• CASTNET
1 SEC Hits
D 2ssn im win
Figure 3-22 U.S. Rural NCore, CASTNET and NPS POMS O3 sites in 2010.
3.5.6.2 Probe/Inlet Siting Requirements
Probe and monitoring path siting criteria for ambient air quality monitoring are
contained in 40 CFR Part 58, Appendix E. For O3, the probe must be located
between 2 and 15 meters above ground level and be at least 1 meter away (both in
the horizontal and vertical directions) from any supporting structure, walls, etc. If it
is located on the side of a building, it must be located on the windward side, relative
to prevailing wind direction during the season of highest potential O3 concentration.
Ozone monitors are placed to determine air quality in larger areas (neighborhood,
urban, or regional scales) and therefore, placement of the monitor probe should not
be near local, minor sources of NO, O3-scavenging hydrocarbons, or O3 precursors.
The probe or inlet must have unrestricted air flow in an arc of at least 180 degrees
and be located away from any building or obstacle at a distance of at least twice the
height of the obstacle. The arc of unrestricted air flow must include the predominant
wind direction for the season of greatest O3 concentrations. Some exceptions can be
made for measurements taken in street canyons or sites where obstruction by
buildings or other structures is unavoidable. The scavenging effect of trees on O3 is
greater than other pollutants and the probe/inlet must be located at least 10 meters
from the tree drip line to minimize interference with normal air flow. When siting O3
monitors near roadways, it is important to minimize the destructive interferences
3-79
-------
from sources of NO, since NO reacts readily with O3. For siting neighborhood and
urban scale O3 monitors, guidance on the minimum distance from the edge of the
nearest traffic lane is based on roadway average daily traffic count (40 CFR Part 58,
Appendix E, Table E-l). The minimum distance from roadways is 10 meters
(average daily traffic count < 1,000) and increases to a maximum distance of
250 meters (average daily traffic count > 110,000).
3.6 Ambient Concentrations
This section investigates spatiotemporal variability in ambient O3 concentrations and
associations between O3 and copollutants. To set the stage for the rest of the section,
common O3 measurement units, metrics, and averaging times are described and
compared in Section 3.6.1. Spatial variability is covered in Section 3.6.2 and is
divided into urban-focused variability and rural-focused variability. Urban-focused
variability is organized by scale, extending from national-scale down to
neighborhood-scale and the near-road environment. Rural-focused variability is
organized by region and includes observations of ground-level vertical O3 gradients
where available. Temporal variability is covered in Section 3.6.3 and is organized by
time, extending from multiyear trends down to hourly (diel) variability. In many
instances, spatial and temporal variability are inseparable (e.g., seasonal dependence
to spatial variability), resulting in some overlap between Section 3.6.2 and
Section 3.6.3. Finally, Section 3.6.4 covers associations between O3 and copollutants
including CO, SO2, NO2, PM2.5, and PM10.
As noted in the 2006 O3 AQCD (U.S. EPA. 2006b). O3 is the only photochemical
oxidant other than nitrogen dioxide (NO2) that is routinely monitored and for which a
comprehensive database exists. Data for other photochemical oxidants (e.g., PAN,
H2O2, etc.) typically have been obtained only as part of special field studies.
Consequently, no data on nationwide patterns of occurrence are available for these
other oxidants; nor are extensive data available on the relationships of concentrations
and patterns of these oxidants to those of O3. As a result, this section focuses solely
on O3, the NAAQS indicator for photochemical oxidants. The majority of ambient
O3 data reported in this section were obtained from AQS, EPA's repository for
detailed, hourly data that has been subject to EPA quality control and assurance
procedures (the AQS network was described in Section 3.5).
3.6.1 Measurement Units, Metrics, and Averaging Times
Several approaches are commonly used for reporting O3 data. In atmospheric
sciences and epidemiology, O3 is frequently reported as a concentration, expressed as
a volume-to-volume mixing ratio, commonly measured in ppm or ppb. In human
exposure, O3 is frequently reported as a cumulative exposure, expressed as a mixing
ratio times time (e.g., ppm-h). In ecology, cumulative exposure indicators are
frequently used that extend over longer time periods, such as growing season or year.
3-80
-------
This section focuses on ambient concentrations derived primarily from hourly
average O3 measurements and concentrations are reported in ppb wherever possible.
Further details on human and ecological exposure metrics can be found in Chapter 4
and Chapter 9, respectively.
As discussed in Section 3.5. most continuous O3 monitors report hourly average
concentrations to AQS with a required precision of 10 ppb and LDL of 10 ppb (see
Table 3-4). This data can be used as reported (1-h avg), or further summarized in one
of several ways to focus on important aspects of the data while simultaneously
reducing the volume of information. Three common daily reporting metrics include:
(1) the average of the hourly observations over a 24-hour period (24-h avg); (2) the
maximum hourly observation occurring in a 24-hour period (1-h daily max); and
(3) the maximum 8-h running average of the hourly observations occurring in a
24-hour period (8-h daily max)1. Throughout this ISA and the literature, O3
concentrations are reported using different averaging times as appropriate, making it
important to recognize the differences between these metrics.
Nation-wide, year-round 1-h avg O3 data reported to AQS from 2007-2009 was used
to compare these different daily metrics. Correlations between the 24-h avg,
1-h daily max and 8-h daily max metrics were generated on a site-by-site basis.
Figure 3-23 contains box plots of the distribution in correlations from all sites.
The top comparison in Figure 3-23 is between 8-h daily max and 1-h daily max O3.
Not surprisingly, these two metrics are very highly correlated (median r = 0.97,
IQR = 0.96-0.98). There are a couple outlying sites, with correlations between these
two metrics as low as 0.63, but 95% of sites have correlations above 0.93.
The middle comparison in Figure 3-23 is between 8-h daily max and 24-h avg O3.
For these metrics, the distribution in correlations is shifted down and broadened out
(median r = 0.89, IQR = 0.86-0.92). Finally, the bottom comparison in Figure 3-23 is
between 1-h daily max and 24-h avg O3. Again, for these metrics the distribution in
correlations is shifted down and broadened out relative to the other two comparisons
(median r = 0.83, IQR = 0.78-0.88). The correlation between the two daily maximum
metrics (1-h daily max and 8-h daily max) are quite high for most sites, but
correlations between the daily maximum metrics and the daily average metric
(24-h avg) are lower. This illustrates the influence of the overnight period on the
24-h avg O3 concentration. In contrast, the 1-h daily max and 8-h daily max are more
indicative of the daytime, higher O3 periods. The correlation between these metrics,
however, can be very site-specific, as is evident from the broad range in correlations
in Figure 3-23 for all three comparisons. Therefore, understanding which O3 metric
is being used in a given study is very important since they capture different aspects
of O3 temporal variability.
1 For O3 regulatory monitoring purposes, the 8-h daily max is calculated by first generating all 8-h running averages and storing
these averages hourly by the first hour in the 8-hour period. The 8-h daily max is then set equal to the maximum of the 24
individual 8-h averages occurring in a given day.
3-81
-------
8-h daily max
vs.
1-h daily max
8-h daily max
vs.
24-h avg
1-h daily max
vs.
24-h avg
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Correlation
Note: Shown are the median (red line), mean (green star), inner-quartile range (box), 5th and 95th percentiles (whiskers), and
extremes (black circles).
Figure 3-23 Distribution in nation-wide year-round site-level correlations
between daily Os metrics including 24-h avg, 1-h daily max and
8-h daily max using AQS data, 2007-2009.
The median 1-h daily max, 8-h daily max, and 24-h avg O3 concentrations across all
sites included in the 3-year nation-wide data set were 44, 40, and 29 ppb,
respectively. Representing the upper end of the distribution, the 99th percentiles of
these same metrics across all sites were 94, 80, and 60 ppb, respectively. While the
ratio of these metrics will vary by location, typically the 1-h daily max will be the
highest value representing peak concentrations and the 24-h avg will be considerably
lower representing daily average concentrations incorporating the overnight period.
The 8-h daily max typically represents the higher mid-day concentrations and will
generally lie somewhere between the other two metrics1.
3.6.2 Spatial Variability
3.6.2.1 Urban-Focused Variability
National-Scale Variability
AQS contains a large depository of national O3 data collected to meet the monitoring
objectives described in Section 3.5.6. In many areas, O3 concentrations decrease
appreciably during months with lower temperatures and decreased sunlight. As a
1 The 8-h daily max is not strictly limited to lie between the 1-h daily max and the 24-h avg since the 8-hour averaging period used to
calculate the 8-h daily max can extend into the morning hours of the subsequent day. However, the 8-h daily max typically
incorporates the middle of the day when O3 concentrations are at their highest, resulting in an 8-h daily max somewhere between
the 1-h daily max and the 24-h avg calculated for that day.
3-82
-------
result, year-round O3 monitoring is only required in certain areas. Table D-3 of 40
CFR Part 58, Appendix D lists the beginning and ending month of the ozone season
(defined in Section 3.5.6.1) by geographic area and Figure 3-24 illustrates these time
periods on a monitor-by-monitor basis. Monitoring is optional outside the ozone
season and many states elect to operate their monitors year-round or for time periods
outside what is strictly mandated.
Required Ozone Monitoring Time Periods
Time Period
• Apr-Sep Mar-Nov
• Apr-Oct • May-Sep
Apr-Nov • May-Oct
Mar-Sep Jun-Sep
• Mar-Oct o Year round
Puerto Rico
Source: U.S. EPA (2008e).
Figure 3-24 Required Os monitoring time periods (ozone season) identified by
monitoring site.
3-83
-------
Hourly FRM and FEM O3 data reported to AQS for the period 2007-2009 were used
to investigate national-scale spatial variability in O3 concentrations. Given the
variability in O3 monitoring time periods available in AQS as a result of the
regionally-varying ozone seasons, the analyses in this section were based on two
distinct data sets:
• a year-round data set: data only from monitors reporting year-round;
• a warm-season data set: data from all monitors reporting May through
September.
The warm-season data set was used to capture the majority of ozone season data
while providing a consistent time-frame for comparison across states. All available
monitoring data including data from year-round monitors was included in the warm-
season data set after removing observations outside the 5-month window. Data were
retrieved from AQS on February 25, 2011 for these two data sets, and all validated
data was included regardless of flags or regional concurrence1. A summary of the
two O3 data sets including the applied completeness criteria is provided in Table 3-5.
Figure 3-25 and Figure 3-26 show the location of the 457 year-round and 1,064
warm-season monitors meeting the completeness criteria for all three years (2007-
2009).
Table 3-5 Summary of O3 data sets originating from AQS.
Years
Months
Completeness Criteria
Year-Round Data Set
2007-2009
January - December (12 mo)
75% of hours in a day
Warm-Season Data Set
2007-2009
May - September (5 mo)
75% of hours in a day
75% of days in a calendar quarter 75% of days between May - September
All 4 quarters per year
Number of monitors meeting
completeness criteria
618 containing at least one valid
year in 2007-2009
1,267 containing at least one valid year in
2007-2009
550 containing at least two valid
years in 2007-2009
1,169 containing at least two valid years in
2007-2009
457 containing all three valid years
in 2007-2009
1,064 containing all three valid years in
2007-2009
1 Concentrations that might have been affected by exceptional events (and contribute to a violation of the NAAQS) can be flagged in
the Air Quality System (AQS) by the reporting organization. Exceptional events are defined as unusual or naturally occurring
events that can affect air quality but are not reasonably controllable using techniques that tribal, state or local air agencies may
implement in order to attain and maintain the National Ambient Air Quality Standards (NAAQS). The corresponding EPA Regional
Office is responsible for reviewing the data and evidence of the event, and deciding whether to concur with the flag. Flagged data
that has been concurred by the Regional office is typically excluded for regulatory purposes.
3-84
-------
Figure 3-25 Location of the 457 Os monitors meeting the year-round data set
completeness criterion for all 3 years between 2007 and 2009.
Figure 3-26 Location of the 1,064 Oz monitors meeting the warm-season data
set completeness criteria for all 3 years between 2007 and 2009.
3-85
-------
Tabulated statistics generated from the year-round and warm-season data sets are
included in Table 3-6 and Table 3-7, respectively. This information was used to
compare (1) the year-round and warm-season data sets; (2) the O3 distribution
variability across years (2005-2009); and (3) four different averaging times (1-h avg,
24-h avg, 1-h daily max, and 8-h daily max). Summary statistics for 2005 and 2006
were added to these tables in order to gain a broader view of year-to-year variability,
but the year-round and warm-season data sets used for analyses in the rest of this
section are limited to 2007-2009 as described above and in Table 3-5.
The 8-h daily max pooled by site was also included in these tables to show the
distribution of the annual and 3-year (2007-2009) site-averages of the 8-h daily max
statistic.
The year-round data set includes data from roughly half the number of monitors as
the warm-season data set and a larger fraction of the year-round monitors are located
in the southern half of the U.S. due to extended monitoring requirements in these
areas. Despite these differences, the mean, SD and percentiles of the nation-wide O3
concentrations were quite similar for the year-round data presented in Table 3-6 and
the warm-season data presented in Table 3-7. In both data sets, there was very little
variability across years in the central statistics; for example, the median 1-h avg
concentrations between 2005 and 2009 ranged from 28 to 29 ppb for the year-round
data and from 29 to 30 ppb for the warm-season data. The 8-h daily max showed
similar uniformity in median across the five years, with concentrations ranging from
39 to 41 ppb for the year-round data and from 40 to 43 for the warm-season data.
The upper percentiles (95th and above) showed a general downward trend from 2005
to 2009 in both nation-wide data sets. For example, the 99th percentile of the
8-h daily max observed in the warm-season data dropped from 85 ppb in 2005 to
75 ppb in 2009. Trends in O3 concentrations investigated over a longer time period
are included in Section 3.6.3.1.
3-86
-------
Table 3-6 Nationwide distributions of O3 concentrations (ppb) from the
year-round data set.
Time N
Period Monitors
NObs
Mean
SD Min 1 5
10 25
50 75
90 95
98 99
Max
Max Site
IDb
1-h avga
2005
2006
2007
2008
2009
2007-2009
499
532
522
520
551
599
4,284,219
4,543,205
4,547,280
4,470,065
4,716,962
13,734,307
29
30
29
30
29
29
18 2
18 2
18 2
17 2
16 2
17 2
2 2
2 2
2 2
2 2
2 2
2 2
2 15
5 16
5 16
6 17
6 17
6 17
28 41
29 42
29 41
29 41
29 40
29 40
53 61
54 61
52 60
52 59
50 56
51 58
71 78
71 78
68 75
67 74
64 70
67 73
182
175
237
222
188
237
06071 0005
060370016
450790021
45021 0002
720770001
450790021
24-h avga
2005
2006
2007
2008
2009
2007-2009
1-h daily max
2005
2006
2007
2008
2009
2007-2009
504
536
531
528
556
611
a
504
536
531
528
556
611
183,815
1 94,884
1 94,873
191,875
202,142
588,890
183,815
1 94,884
1 94,873
191,875
202,142
588,890
29
30
29
30
29
29
48
48
47
47
45
46
13 2
13 2
12 2
12 2
11 2
12 2
18 2
18 2
17 2
17 2
15 2
16 2
4 9
5 10
5 11
5 11
6 11
5 11
11 21
13 23
14 23
14 23
14 22
14 23
13 20
14 21
14 20
14 21
14 21
14 21
26 35
28 36
28 36
27 35
27 35
27 35
28 37
29 38
29 37
29 38
28 37
29 37
46 58
46 58
45 57
45 56
44 54
44 55
46 51
47 52
45 50
46 50
44 48
45 49
71 80
71 80
69 77
67 76
64 72
67 75
57 61
58 62
56 60
56 61
53 57
55 60
91 100
91 100
87 94
87 96
83 91
86 94
103
102
96
98
95
98
182
175
237
222
188
237
06071 9002
061 070009
060651016
06071 0005
06071 0005
06071 0005
06071 0005
060370016
450790021
45021 0002
720770001
450790021
8-h daily max'1
2005
2006
2007
2008
2009
2007-2009
8-h daily max
2005
2006
2007
2008
2009
2007-2009
504
536
528
528
556
608
(pooled by
508
538
538
529
558
457
183,279
1 94,285
1 94,266
191,283
201 ,536
587,085
site)"
508
538
538
529
558
457
42
42
41
41
40
41
42
42
41
41
40
41
"Includes all validated data regardless of flags or
purposes
bAQS Site ID corresponding to the observation in
16 2
16 2
15 2
15 2
14 2
15 2
6 23
6 12
6 17
6 20
6 20
6 19
7 16
9 18
10 19
11 19
11 18
10 19
27 32
28 31
27 31
28 31
26 30
29 32
21 30
23 31
23 31
23 31
23 30
23 31
34 38
34 38
34 38
34 37
33 36
34 38
40 52
41 52
40 51
40 51
39 49
40 50
42 45
43 46
41 45
40 45
39 44
40 45
63 70
63 70
61 68
60 66
57 63
60 66
48 51
50 52
49 51
50 52
48 50
49 51
78 84
79 85
75 81
75 82
71 77
74 80
53 55
54 55
54 55
55 57
53 54
54 55
145
142
137
172
128
172
61
61
63
61
60
61
06071 0005
06071 0005
06071 0005
45021 0002
06071 2002
45021 0002
06071 0005
06071 9002
06071 9002
06071 9002
06071 9002
06071 9002
regional concurrence and therefore may differ from data used for regulatory
the Max column
3-87
-------
Table 3-7 Nationwide distributions of O3 concentrations (ppb) from the
warm-season data set.
Time
Period
N
Monitors
NObs
Mean
SD
Min
1
5
10
25
50
75
90
95
98
99
Max
Max Site
IDb
1-havga
2005
2006
2007
2008
2009
2007-2009
1,023
1,036
1,021
1,034
1,029
1,103
7,455,018
7,590,796
7,711,463
7,701 ,597
7,835,074
23,248,134
30
31
31
31
29
30
19
18
18
17
16
17
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
5
6
6
7
7
7
16
17
18
18
17
18
29
30
30
30
29
30
43
43
43
42
40
42
55
55
55
53
50
53
64
62
63
60
56
60
73
71
71
68
63
68
79
77
77
74
69
74
182
175
237
222
259
259
06071 0005
060370016
450790021
45021 0002
311090016
311090016
24-h avga
2005
2006
2007
2008
2009
2007-2009
1,103
1,110
1,100
1,120
1,141
1,197
319,410
324,993
330,197
329,918
335,669
995,784
30
31
31
31
29
30
12
12
12
12
11
12
2
2
2
2
2
2
5
6
6
6
6
6
10
12
12
12
12
12
14
15
16
16
15
16
22
22
23
22
21
22
30
30
31
30
29
30
39
39
39
38
37
38
46
47
47
46
44
45
51
52
51
50
48
50
57
58
57
56
53
55
61
61
61
60
56
59
103
102
96
98
95
98
06071 9002
061 070009
060651016
06071 0005
06071 0005
06071 0005
1-h daily max'1
2005
2006
2007
2008
2009
2007-2009
8-h daily max
2005
2006
2007
2008
2009
2007-2009
8-h daily max
2005
2006
2007
2008
2009
2007-2009
1,103
1,110
1,100
1,120
1,141
1,197
a
1,104
1,112
1,097
1,120
1,141
1,194
(pooled by
1,141
1,152
1,164
1,163
1,173
1,064
319,410
324,993
330,197
329,918
335,669
995,784
318,771
324,327
329,482
329,223
334,972
993,677
site)"
1,141
1,152
1,164
1,163
1,173
1,064
50
50
50
48
46
48
44
44
44
43
40
42
45
44
45
43
41
43
18
17
17
16
15
16
16
16
15
15
13
15
6
6
7
6
5
6
2
2
2
2
2
2
2
2
2
2
2
2
14
12
17
20
20
19
12
15
16
16
15
16
9
11
12
12
12
12
28
29
28
29
28
29
23
25
25
25
23
24
18
20
20
20
19
20
34
34
34
33
32
34
28
29
30
29
28
29
23
25
25
25
24
24
36
37
36
36
35
36
38
38
38
37
36
37
32
33
33
33
31
32
41
41
40
39
38
39
49
48
48
47
45
47
43
43
43
42
40
42
46
45
45
44
41
43
61
60
60
58
54
58
55
54
54
52
49
52
49
48
50
48
44
47
74
72
72
69
64
68
66
64
65
61
57
61
52
51
54
50
47
50
81
80
80
76
71
76
72
70
71
67
63
67
54
54
56
53
50
52
91
90
88
86
80
85
79
78
78
74
69
75
56
58
58
56
53
55
99
98
95
93
87
93
85
84
82
80
75
80
57
59
59
58
55
57
182
175
237
222
259
259
145
142
137
172
128
172
61
65
64
61
63
61
06071 0005
060370016
450790021
45021 0002
311090016
311090016
06071 0005
06071 0005
06071 0005
45021 0002
06071 2002
45021 0002
040139508
060170020
471 5501 02
06071 9002
060651016
06071 9002
"Includes all validated data regardless of flags or regional concurrence and therefore may differ from data used for regulatory
purposes.
bAQS Site ID corresponding to the observation in the Max column.
-------
Given the strong diurnal pattern in O3 concentrations, the selection of averaging time
has a substantial effect on the magnitude of concentration reporting. The nation-wide
median 1-h avg, 24-h avg, 1-h daily max, and 8-h daily max concentrations for the
year-round data set in 2009 were 29, 28, 44 and 39 ppb, respectively. The median
concentrations for the warm-season data set in 2009 were: 29, 29, 45 and 40 ppb,
respectively. The 1-h avg and 24-h avg both include the lowest concentrations
typically observed in the overnight period which lowers their values relative to the
daily maximum statistics.
A strong seasonal pattern in O3 concentrations can also be seen in the year-round
data. Table 3-8 shows the 8-h daily max stratified by season, with the seasons
defined as:
• winter: December-February;
• spring: March-May;
• summer: June-August;
• fall: September-November.
In addition, warm-season (May-Sept) and cold-season (Oct-Apr) stratifications of the
year-round data set are included in the table for comparison with the four seasonal
stratifications. Substantial seasonal variability in the 8-h daily max concentration for
the period 2007-2009 was evident with lower concentrations present in fall
(median = 36 ppb) and winter (median = 32 ppb) and higher concentrations in spring
(median = 47 ppb) and summer (median = 46 ppb). The seasonal differences were
even more pronounced in the upper percentiles. For example, the 99th percentile in
the 8-h daily max over the 2007-09 time period ranged from 52 ppb in winter to
90 ppb in summer. The distribution in 8-h daily max O3 during the warm-season (as
defined above) and during summer were very similar, which is not surprising given
their close overlap in months. The distribution during the cold-season (as defined
above) is shifted toward higher 8-h daily max O3 concentrations compared with the
distribution during winter. This is a result of including the four transition months
(Oct, Nov, Mar and Apr) in the cold-season when high O3 concentrations can occur.
Further investigation of temporal variability including multiyear trends and diel
behavior is included in Section 3.6.3.
3-89
-------
Table 3-8 Seasonally stratified distributions of 8-h daily max O3
concentrations (ppb) from the year-round data set (2007-2009).
Max Site
b
Time Period
Monitors N Obs Mean
SD
Min
1
5
10
25
50
75
90
95
98
99
Max
ID
8-h daily max (2007-2009)3
Year-round
8-h daily max by
Winter (Dec-Feb)
Spring (Mar-May)
Summer
(June-Aug)
Fall (Sept-Nov)
Warm-season
(May-Sept)
Cold-season
(Oct-Apr)
608
season
608
612
613
608
616
608
587,085
(2007-2009)3
143,855
148,409
148,280
146,541
246,233
340,852
41
31
47
47
37
47
36
15
10
12
16
13
16
12
2
2
2
2
2
2
2
10
6
20
16
10
16
8
19
14
28
22
17
22
16
23
18
33
26
21
27
21
31
25
40
35
28
35
28
40
32
47
46
36
46
36
50
38
55
57
45
57
44
60
43
62
67
54
66
52
66
46
67
75
61
73
57
74
49
72
84
68
81
63
80
52
77
90
75
87
67
172
172
118
137
116
137
172
450210002
450210002
06037001 6
060710005
06037001 6
060710005
450210002
"Includes all validated data regardless of flags or regional concurrence and therefore may differ from data used for regulatory
purposes.
bAQS Site ID corresponding to the observation in the Max column.
A national picture of AQS O3 concentrations was generated from the year-round and
warm-season data sets by aggregating the 8-h daily max observations by U.S. county.
For this purpose, the 8-h daily max concentrations at each site were averaged over
one or more calendar years and then the highest site in each county was selected for
that county. Figure 3-27 contains the county-scale 8-h daily max O3 concentrations
from the year-round data set for 2007-2009 (top map) with seasonal stratification
(bottom four maps). Figure 3-28 contains the county-scale 8-h daily max O3
concentrations from the warm-season data set for 2007-2009 (top map) along with
individual maps for each calendar year between 2007 and 2009 (bottom three maps).
These maps are meant to illustrate the general national-scale distribution in long-term
average 8-h daily max O3 concentrations and are not representative of O3
concentrations at all locations or times within the counties shown; considerable
spatial variability can exist within a county. This is particularly important in the West
where counties are larger on average than in the East. These maps are limited by
monitor availability, resulting in the majority of U.S. counties not having available
data (the white regions in Figure 3-27 and Figure 3-28).
3-90
-------
O3 > 60 ppb
50 < O3 < 60 ppb
40 < O3 < 50 ppb
30 < O3 < 40 ppb
O3 < 30 ppb
Winter
Spring
Figure 3-27 Highest monitor (by county) 3-year avg (2007-2009) of the
8-h daily max O3 concentration based on the year-round data set
(the top map) with seasonal stratification (the four bottom maps).
3-91
-------
O3 > 60 ppb
50 < O3 < 60 ppb
40
-------
As shown in the top county-scale map generated from the 2007-2009 year-round data
set in Figure 3-27, the highest 3-year avg 8-h daily max O3 concentrations (> 50 ppb)
occur in counties in central and southern California, Arizona, Colorado and high
elevation counties in Tennessee. The highest year-round average concentration of
61 ppb over this period comes from Site #060719002 located at an elevation of
1,244 meters in Joshua Tree National Monument, San Bernardino County, CA.
The lowest 3-year avg 8-h daily max O3 concentrations (<30 ppb) occur in Pacific
Coast counties in northern California and Washington as well as in two northeastern
counties in Pennsylvania and Massachusetts. The seasonally-stratified county-scale
maps in Figure 3-28 reinforce the strong seasonality in 8-h daily max O3
concentrations shown in Table 3-8. The highest wintertime concentrations (> 40 ppb)
occur in the West with the highest 3-year wintertime avg of 46 ppb calculated for
Site #080690007 located at an elevation of 2,743 meters near Rocky Mountain
National Park, Larimer County, Colorado. In spring and summer, the concentrations
increase considerably across all counties, with the highest concentrations (> 60 ppb)
occurring during the summer in 15 counties in California, 3 counties in Colorado and
1 county each in Nevada and Arizona. Many counties in rural Wyoming, Montana,
North Dakota, Maine, and along the Gulf Coast peak in the spring instead of the
summer. In the fall, 8-h daily max O3 concentrations drop back down below their
spring and summer concentrations.
The top county-scale map in Figure 3-28 based on the 2007-2009 warm-season data
set looks similar to the corresponding map in Figure 3-27 based on the year-round
data set. The warm-season map, however, incorporates approximately twice as many
monitors across the U.S., providing more spatial coverage. Several counties in Utah,
New Mexico, Indiana, Ohio, Maryland, North Carolina, and Georgia in addition to
California, Arizona, Colorado and Tennessee identified above have 3-year avg
(2007-2009) 8-h daily max O3 concentrations > 50 ppb based on the warm-season
data set. The individual yearly average county-maximum 8-h daily max O3
concentrations in the lower half of Figure 3-27 show a general decrease in most
counties from 2007 to 2009. The number of counties containing a monitor reporting
an annual average 8-h daily max O3 concentration above 50 ppb dropped from 230
counties in 2007 to 30 counties in 2009. This is consistent with the general decrease
across these years shown in Table 3-6 and Table 3-7 for the upper percentiles of the
8-h daily max O3 concentration.
Urban-Scale Variability
Statistical analysis of the human health effects of airborne pollutants based on
aggregate population time-series data have often relied on ambient concentrations of
pollutants measured at one or more central monitoring sites in a given metropolitan
area. The validity of relying on central monitoring sites is strongly dependent on the
spatial variability in concentrations within a given metropolitan area. To investigate
urban-scale variability, 20 focus cities were selected for closer analysis of O3
concentration variability; these cities are listed in Table 3-9 and were selected based
on their importance in O3 epidemiology studies and on their geographic distribution
3-93
-------
across the United States. In order to provide a well-defined boundary around each
city, the combined statistical area (CSA) encompassing each city was used. If the city
was not within a CSA, the smaller core-based statistical area (CBSA) was selected.
The CSAs/CBSAs are defined by the U.S. Census Bureau (20II)1 and have been
used to establish analysis regions around cities in previous ISAs for particulate
matter (PM) (U.S. EPA. 2009d) and carbon monoxide (CO) (U.S. EPA. 2010c).
The distribution of the 8-h daily max O3 concentrations from 2007-2009 for each of
the 20 focus cities is included in Table 3-10. These city-specific distributions were
extracted from the warm-season data set and can be compared to the nationwide
warm-season 8-h daily max distribution for 2007-2009 in Table 3-7 (and repeated in
the first line of Table 3-10 for reference). The median 8-h daily max concentration in
these focus cities was 41 ppb, similar to the nationwide median of 42 ppb. Seattle
had the lowest median (31 ppb) and Salt Lake City had the highest median (53 ppb)
of the 20 cities investigated. The 99th percentile of the 8-h daily max concentration
in the focus cities was 84 ppb; similar once again to the nationwide 99th percentile of
80 ppb. Seattle had the lowest 99th percentile (64 ppb) and Los Angeles had the
highest 99th percentile (98 ppb) of the 20 cities investigated. In aggregate, the 20
focus cities selected are similar in distribution to the nationwide data set, but there is
substantial city-to-city variability in the individual distributions of the 8-h daily max
concentrations based on the warm-season data set.
Maps showing the location of central monitoring sites with O3 monitors reporting to
AQS for each of the 20 focus cities are included as supplemental material in
Section 3.9.L Figure 3-76 through Figure 3-95: examples for Atlanta,GA, Boston,
MA, and Los Angeles, CA, are shown in Figure 3-29 through Figure 3-31. The sites
are delineated in the maps as year-round or warm-season based on their inclusion in
the year-round data set and the warm-season data set (the warm-season data set
includes May-September data from both the warm-season monitors and the year-
round monitors meeting the warm-season data inclusion criteria). The maps also
include the CSA/CBSA boundary selected for monitor inclusion, the location of
urban areas and water bodies, the major roadway network, as well as the population
gravity center based on the entire CSA/CBSA and the individual focus city
boundaries. Population gravity center is calculated from the average longitude and
latitude values for the input census tract centroids and represents the mean center of
the population in a given area. Census tract centroids are weighted by their
population during this calculation.
1A CBSA represents a county-based region surrounding an urban center of at least 10,000 people determined using 2000 census
data and replaces the older Metropolitan Statistical Area (MSA) definition from 1990. The CSA represents an aggregate of
adjacent CBSAs tied by specific commuting behaviors. The broader CSA definition was used when selecting monitors for the
cities listed above with the exception of Phoenix and San Antonio, which are not contained within a CSA. Therefore, the smaller
CBSA definition was used for these metropolitan areas.
3-94
-------
Table 3-9 Focus cities used in this and previous assessments.
Focus City
Atlanta, GA
Baltimore, MD
Birmingham, AL
Boston, MA
Chicago, IL
Dallas, TX
Denver, CO
Detroit, Ml
Houston, TX
Los Angeles, CA
Minneapolis, MN
New York, NY
Philadelphia, PA
Phoenix, AZ
Pittsburgh, PA
Salt Lake City,
UT
San Antonio, TX
San Francisco,
CA
Seattle, WA
St Louis, MO
Short Name
Atlanta CSA
Baltimore CSA
Birmingham CSA
Boston CSA
Chicago CSA
Dallas CSA
Denver CSA
Detroit CSA
Houston CSA
Los Angeles CSA
Minneapolis CSA
New York CSA
Philadelphia CSA
Phoenix CBSA
Pittsburgh CSA
Salt Lake City
CSA
San Antonio
CBSA
San Francisco
CSA
Seattle CSA
St Louis CSA
"Defined based on 2000 Census data
Year-Round O3
CSA/CBSA Name3 Monitoring Sites'3
Atlanta-Sandy
Springs-Gainesville
Washington-
Baltimore-northern VA
Birmingham-Hoover-
Cullman
Boston-Worcester-
Manchester
Chicago-Nape rville-
Michigan City
Dallas-Fort Worth
Denver-Aurora-
Boulder
Detroit-Warren-Flint
Houston-Baytown-
Huntsville
Los Angeles-Long
Beach-Riverside
Minneapolis-St. Paul-
St. Cloud
New York-Newark-
Bridgeport
Philadelphia-Camden-
Vineland
Phoenix-Mesa-
Scottsdale
Pittsburgh-New Castle
Salt Lake City-Ogden-
Clearfield
San Antonio
San Jose-
San Francisco-
Oakland
Seattle-Tacoma-
Olympia
St. Louis-St. Charles-
Farmington
from the U.S. Census Bureau
0
9
1
3
11
19
12
0
21
47
2
20
9
14
2
2
5
25
5
3
(2011).
Warm-Season O3 Included in
Monitoring Sites0 Prior ISAsd
11
19
9
18
15
0
3
9
0
3
6
10
8
17
12
10
0
6
5
13
CO, PM, SOX,
NOX
NOX
PM
CO, PM, NOX
PM, NOX
CO, PM
PM
CO, PM, NOX
CO, PM, SOX,
NOX
CO, PM, SOX,
NOX
PM, NOX
CO, PM
CO, PM
CO, PM
CO, PM, SOX
bThe number of sites within each CSA/CBSA with AQS monitors meeting the year-round data set inclusion criteria.
°The number of sites within each CSA/CBSA with AQS monitors meeting the warm-season data set inclusion criteria; the warm-
season data set includes May - September data from both the warm-season and year-round monitors meeting the warm-season
data set inclusion criteria.
"Boundaries for the 2010 CO ISA (U.S. EPA. 201 Oc) and 2009 PM ISA (U.S. EPA. 2009d) focus cities were based on CSA/CBSA
definitions; boundaries for the 2008 SOX ISA (U.S. EPA. 2008d) and 2008 NOX ISA (U.S. EPA. 2008c) focus cities were based
on similar metropolitan statistical area (MSA) definitions from the 1990 U.S. Census.
3-95
-------
Table 3-10 City-specific distributions of 8-h daily max O3 concentrations (ppb)
from the warm-season data set (2007-2009).
Time Period
8-h daily max
Nationwide
8-h daily max
Atlanta, GA,
CSA
Baltimore, MD,
CSA
Birmingham,
AL, CSA
Boston, MA,
CSA
Chicago, IL,
CSA
Dallas, TX,
CSA
Denver, CO,
CSA
Detroit, Ml,
CSA
Houston, TX,
CSA
Los Angeles,
CA, CSA
Minneapolis,
MN.CSA
New York, NY,
CSA
Philadelphia,
PA, CSA
Phoenix, AZ,
CBSA
Pittsburgh, PA,
CSA
Salt Lake City,
UT, CSA
San Antonio,
TX, CSA
San Francisco,
CA, CSA
Seattle, WA,
CSA
St Louis, MO,
CSA
All
CSAs/CBSAs
listed
N
Monitors N Obs
Mean SD
Min
1 5
10
25
50
75
90
95
98
99
Max
Max Site
IDb
(2007-2009)3
1,194
993,677
42
15
2
12 20
24
32
42
52
61
67
75
80
172
450210002
by CSA/CBSA (2007-2009)3
11
28
10
21
27
19
15
9
21
49
8
21
14
22
13
12
5
31
5
19
360
7,844
20,999
7,676
12,603
20,764
19,858
12,217
5,016
22,305
49,295
5,315
26,304
12,673
26,129
9,814
5,146
4,701
28,325
6,148
1 1 ,569
314,701
47
43
44
41
37
41
44
45
36
47
40
39
41
49
43
51
39
34
31
43
42
16
16
15
14
14
15
15
14
15
18
12
16
17
12
15
14
13
12
12
15
16
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
15 22
9 18
14 21
13 21
9 15
11 20
8 18
15 23
8 15
10 20
15 21
6 15
8 17
18 27
12 19
8 23
13 20
8 16
4 12
12 19
9 18
27
23
25
25
19
24
24
28
19
26
25
20
21
32
24
32
23
20
17
23
22
36
31
34
31
27
31
34
35
25
35
31
28
29
41
32
44
29
26
23
32
31
47
43
44
40
37
39
44
44
34
45
40
37
39
50
43
53
37
33
31
43
41
58
54
54
49
47
50
55
52
46
58
48
47
52
58
53
61
46
41
39
53
52
67
64
63
59
57
61
63
62
57
72
54
59
64
65
62
67
56
48
46
61
63
72
70
68
67
62
67
68
69
64
81
58
68
70
68
68
71
62
55
51
68
69
81
78
76
75
69
74
72
77
72
91
63
77
78
72
74
77
67
63
59
76
78
87
83
83
81
74
79
76
83
78
98
67
83
83
75
78
80
72
68
64
81
84
124
118
108
104
108
121
98
100
110
137
86
123
125
85
100
96
90
110
91
113
137
130890002
24003001 4
01 0732006
25027001 5
170310042
484390075
080590006
260990009
482011034
060710005
270031 002
090050005
2401 50003
040137021
420050001
490353008
480290032
060010007
530330023
2951 00086
060710005
"Includes all validated data regardless of flags or regional concurrence and therefore may differ from data used for regulatory
purposes.
bAQS Site ID corresponding to the observation in the Max column.
3-96
-------
Legend
Monitor Location*
O Warm-Mason Monitor*
• Veer-round
• City-twwwJ Population Gravity C*f*r
Figure 3-29 Map of the Atlanta, Georgia, CSA including Os monitor locations,
population gravity centers, urban areas, and major roadways.
Legend
Monitor Location.
O Warm-season Monitors
• Year-round Monitor*
• Cffy-bated Population Gravity C*n«f
• CSA-bMAd Population Gravity Cental
Interstate Highways
Mapy Highway*
Figure 3-30 Map of the Boston, Massachusetts, CSA including O$ monitor
locations, population gravity centers, urban areas, and major
roadways.
3-97
-------
Legend
Monitor Locations
O Vl&rm-season Monitors
• Year-round Monitors
• City-based Population Cravity Center
• CSA-based Population Gravity Center
Interstate Highways
Major Highways
Water Bod.es
Urban Areas
Los Angeles CSA
.
200 Kilomelers
Figure 3-31 Map of the Los Angeles, California, CSA including O3 monitor
locations, population gravity centers, urban areas, and major
roadways.
The Atlanta, GA, CSA contains 11 warm-season monitors distributed evenly yet
sparsely around the city center (Figure 3-29). The population gravity center for the
city and the larger CSA are only separated by 4 km, indicating that the majority of
the population lives within or evenly distributed around the city limits. Atlanta is
landlocked with a radial network of interstate highways leading to the city center.
The Boston, MA, CSA contains 3-year-round and 18 warm-season monitors spread
evenly throughout the CSA. Boston is a harbor city with the Atlantic Ocean to the
east, resulting in the city-based population gravity center being located 17 km east of
the CSA-based population gravity center. The Los Angeles, CA, CSA contains the
largest number of monitors of the 20 CSA/CBSAs investigated with 47 year-round
and 3 warm-season monitors. These monitors are primarily concentrated in the
Los Angeles urban area with relatively few monitors extending out to the northern
and eastern reaches of the CSA. These unmonitored areas are very sparsely
populated, resulting in only 15 km separating the city-based and the CSA-based
population gravity centers despite the vast area of the Los Angeles CSA.
Other CSAs/CBSAs (see Section 3.9.1) with monitors concentrated within the focus
city limits include Birmingham, AL, Chicago, IL, Denver, CO, Houston, TX,
Phoenix AZ, San Antonio, TX, and Salt Lake City, UT. The remaining CSAs/CBSAs
have monitors distributed more evenly throughout the CSA/CBSA area. Baltimore,
MD, is contained within the same CSA as Washington D.C. and suburbs, resulting in
3-98
-------
a 50-km separation (the largest of the focus cities investigated) between the city-
based population gravity center for Baltimore and the CSA-based population gravity
center for the Washington-Baltimore-Northern Virginia CSA.
Box plots depicting the distribution of 2007-2009 warm-season 8-h daily max O3
data from each individual monitor in the 20 focus cities are included as supplemental
material in Section 3.9.2, Figure 3-96 through Figure 3-115; examples for Atlanta,
GA, Boston, MA, and Los Angeles, CA, are shown in Figure 3-32 through
Figure 3-34. The Atlanta CSA has little spatial variability in 8-h daily max O3
concentrations with median concentrations ranging from 47 ppb at Sites I and J
located far from the city center to 54 ppb at Site A located closest to the city center.
The variation in warm-season 8-h daily max concentrations are also relatively similar
across monitors with IQRs ranging from 17 ppb at Site J to 23 ppb at Site B.
The Boston CSA has more spatial variability in 8-h daily max O3 concentrations than
the Atlanta CSA with median concentrations ranging from 33 ppb at Site A nearest to
the city center to 46 ppb at Site L located 84 km west of the city center. For monitors
located within and just adjacent to the Boston city limits (Sites A-D), the O3
concentrations can vary over relatively short distances owing to differing degrees of
NOX titration and influence from the local topography. Like the Atlanta CSA, the
variation in warm-season 8-h daily max concentrations are relatively similar across
monitors within the Boston CSA with IQRs ranging from 15 ppb at Site U to 21 ppb
at Site K. The Los Angeles CSA exhibits the most variability in O3 concentrations
between monitors of all the CSAs/CBSAs investigated. The median 8-h daily max
O3 concentration in the Los Angeles CSA ranged from 20 ppb at Site AM in the
south-central extreme of the CSA to 80 ppb at Site AE near Crestline, CA in the San
Bernardino National Forest just north of San Bernardino, CA. These two sites are at
approximately the same longitude and are separated by only 85 km, but the Crestline
site is downwind of the Los Angeles basin, resulting in substantially higher O3
concentrations. Site AM also contains data for only 2009, which could explain some
of the deviation when comparing this site with others in the Los Angeles CSA. Sites
AM and AE also had the lowest (8 ppb) and highest (28 ppb) IQR, respectively.
The remaining focus cities exhibited spatial variability ranging from uniform as in
the Atlanta CSA to non-uniform as observed in the Los Angeles CSA (see
supplemental figures in Section 3.9.2).
3-99
-------
Site ID
131210055
130890002
131350002
130670003
132470001
130970004
131130001
131510002
130770002
130850001
132230003
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-08
07-09
07-09
07-09
07-09
Key
Ł>r> f
N
450
452
446
459
450
455
306
459
455
458
455
ra
3 Ł
9
Mean
53
52
52
51
51
52
52
51
47
47
50
c
3
T3
E
1
SD
17
18
16
16
18
15
15
17
16
13
14
iC
76'^
1
Atlanta CSA
Median IQR Site .... i ,,., i ....
54 22
52 23
52 18
52 22
51 22
53 22
52 20
51 22
47 19
47 17
50 21
r-
r"
"ib
o>
--H
A-
B-
c-
D-
E-
F-
G-
H-
I-
J-
K-
f i__d— — i" — *
rHi
-JEEj- 1 i
\—r3
[ — -\
[ 4
' '
~\- ^
] H
-A
-B
-c
-D
-E
-F
-G
-H
-I
-J
-K
.... I .... I ....
0 50 100 150
03 (ppb)
Figure 3-32 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Atlanta CSA.
Boston CSA
Site ID
250250042
250250041
250092006
250213003
250171102
250170009
250095005
330111011
250270024
250094004
440071010
250270015
330110020
330150016
330115001
330150014
250051002
440030002
330131007
440090007
330012004
Years
07-09
08-09
07-09
07-09
07-09
07-09
07-09
07-09
09
07-08
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
iO
N
459
306
459
459
457
439
459
457
153
305
453
458
455
458
459
459
459
458
459
459
459
c
ra
*ib
[--
*ur>
o>
-H
A-
B-
c-
D-
E-
F-
G-
H-
I-
J-
K-
L-
M-
N-
0-
p-
Q-
R-
s-
T-
U-
(
, , , , i , , , , i , , , ,
>--! j* JI- — ^
i---| ^ ]--i
I---I ;|« I--- -H
E- --\ t* 1 1
"" " "| j |> ~|" ----- H
[- - -| |t ] [
'" " 1 ! fr ^ ^
r !6 1 H
J- - -| |>j ~|- i
-A
-B
-c
-D
-E
-F
-G
-H
-I
-J
-K
-L
-M
-N
-0
-P
-Q
-R
-S
-T
-U
) 50 100 150
03 {ppb)
Figure 3-33 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Boston CSA.
3-100
-------
Los Angeles CSA
Site ID
060371602
060371301
060371302
060371103
060372005
060374002
060595001
060590007
060375005
060371002
060370002
060370113
060370016
060371701
060591 003
060371201
060711004
060376012
060650004
060592022
061112002
060658005
060712002
060658001
061110007
060710012
060379033
061110009
060719004
060659001
060710005
060656001
060714003
060714001
060710306
061113001
061111004
061112003
060650009
060650012
060651016
060710001
060655001
060719002
060652002
060651999
060651010
060711234
060650008
060659003
Years
07-09
07-08
09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
08
07-09
07-09
08-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
09
07-09
07-09
07-09
07-09
07-09
07-09
07-08
09
07-09
07,09
07-09
Key
« I
N
458
306
152
457
459
459
459
459
459
459
459
459
458
459
459
459
457
457
127
457
455
276
459
440
459
456
452
458
457
453
459
459
459
455
459
453
458
457
153
457
459
455
459
452
448
283
153
453
265
444
C
(13
! 1
«
Mean
48
36
44
46
54
38
50
48
45
56
57
48
64
61
45
61
66
68
69
52
62
65
68
69
54
67
67
58
70
68
79
72
73
68
64
44
57
41
22
73
73
61
69
73
62
49
59
59
58
42
tz
fl
1
1
SD
13
9
10
12
15
10
12
10
9
14
17
10
18
16
9
14
19
18
18
13
12
15
19
16
10
13
13
11
19
16
19
17
18
14
12
9
11
9
8
15
16
11
14
13
13
17
10
10
10
10
1
I
Median IQR Site
47 17
34 10
44 12
45 14
53 18
37 11
49 14
47 12
45 12
55 19
56 22
47 13
63 23
60 20
44 12
60 19
66 23
69 27
65 23
50 15
62 16
64 18
67 24
68 18
54 12
67 18
66 19
58 14
70 26
67 21
80 28
73 24
73 25
68 21
64 17
43 11
57 14
40 12
20 8
71 22
73 23
60 15
68 21
73 18
61 18
50 22
59 15
58 13
57 14
42 13
1
1 —
CT>
A-
B-
c-
D-
E-
F-
G-
H-
1-
J-
K-
L-
M-
N-
0-
p
Q-
R-
s-
T-
u-
V-
w-
X-
Y-
z-
AA-
AB-
AC-
AD-
AE-
AF-
AG-
AH-
Al-
AJ-
AK-
AL-
AM-
AN-
AO-
AP-
AQ-
AR-
AS-
AT-
AU-
AV-
AW-
AX-
(
i , , , , i
>---! K Ir— ^
J- - [ fr "j- - - -* |
I- 1 fr ] - 1 - - -;
3 \ ^', "] 1
V • -| f \' • M |
r - - -j^ Jit |! H
;- - -| j» f * - - -j
J-4-JU *"
J--PT1--M
^ j ;• *. ' tH
j. [ 1 > [- !
• • > — G~^ i — "*
i- I! E h ^
; j ^
13SK]* ~ • 1 V > r — H
1 T, 1
J, - , -iT m 1- - - - J|
|- [ fr ~| H
**" " 4 9 l~" " ^
\ / • 1
*• '14 1 "* ^
j. -._ -r 1 j L ._ _ - ^
>-L»J--" <=
*" JA 4 i " '
j--nn--4
.-..[-pi.--.
:---ryi---4
!- - -[-| )» | 1
*"""* "1 f 1 "*
!- - -j | |> | H
[- 1| fr ~[ H
!••!--[ f \ ^
^---i i t I-----?
i j ii j
. ! 1
*" — | y ] — i
-A
-B
-c
-D
-E
-F
-G
-H
-I
- J
-K
-L
-M
-N
-o
-P
-Q
-R
-s
-T
-u
-V
-w
-x
-Y
-Z
-AA
-AB
-AC
-AD
-AE
-AF
-AG
-AH
-Al
-AJ
-AK
-AL
-AM
-AN
-AO
-AP
-AQ
-AR
-AS
-AT
-AU
-AV
-AW
-AX
) 50 100 150
03 (ppb)
Figure 3-34 Site information, statistics and box plots for 8-h daily max Os from
AQS monitors meeting the warm-season data set inclusion criteria
within the Los Angeles CSA.
3-101
-------
Pair-wise monitor comparisons were used to further evaluate spatial variability
between monitors within the 20 focus cities. In the particular case of ground-level
O3, central-site monitoring has been justified as a regional measure of exposure
mainly on the grounds that correlations between concentrations at neighboring sites
measured over time are usually high. In areas with multiple monitoring sites,
averages over the monitors have often been used to characterize population
exposures. However, substantial differences in concentrations between monitors can
exist even though concentrations measured at the monitoring sites are highly
correlated, thus leading to the potential for exposure misclassification error.
Therefore, both the Pearson correlation coefficient and the coefficient of divergence
(COD) were calculated for each monitor pair within the CSA/CBSAs using the
8-h daily max O3 data. The correlation provides an indication of temporal linear
dependence across sites while the COD provides an indication of the variability in
absolute concentrations across sites. The COD is defined as follows:
CODik= ^/*'/-**
jk ,,— \-i
•\| '-1
Equation 3-1
where JG/ and Xik represent observed concentrations averaged over some
measurement averaging period / (hourly, daily, etc.) at sites j and k, and/? is the
number of paired observations. A COD of 0 indicates there are no differences
between concentrations at paired sites (spatial homogeneity), while a COD
approaching 1 indicates extreme spatial heterogeneity. These methods for analysis of
spatial variability follow those used in previous ISAs for CO, PM, SOX and NOX as
well as those used in Pinto et al. (2004) for PM2.5.
Histograms and contour matrices of the Pearson correlation coefficient between
8-h daily max O3 concentrations from each monitor pair are included as
supplemental material in Section 3.9.3. Figure 3-116 through Figure 3-135: examples
for Atlanta, Boston and Los Angeles are shown in Figure 3-35 through Figure 3-37.
Likewise, histograms, contour matrices, and scatter plots of the coefficient of
divergence (COD) between 8-h daily max O3 concentrations from each monitor pair
are included as supplemental material in Section 3.9.3. Figure 3-136 through
Figure 3-155: examples for Atlanta, Boston and Los Angeles are shown in
Figure 3-38 through Figure 3-40. These figures also contain scatter plots of
correlation and COD as a function of straight-line distance between monitor pairs.
3-102
-------
Atlanta CSA
20-
15-
10
5-
-0.1 0.0 0.1 0.2 0.3
0.4 0.5
Correlation
co o
1.0-
0.9-
08
0.7-
0.6-
c
I 0.5-
° 0.4-
0.3-
0.2
O.H
00
-0.1
4 *
*•.>;'.*
0.75 076
0.82 088 0.90 0.87 074 075
0.77 0.73 0.75 0,78 0.79 0.68
0.90 0.82 0.77 0.81 081
085 058
084 076 088 075
061 076
0.86 0,63 070 I H
069
0.81
50 100 150 200 250 300 350 400 450
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of the correlations.
Figure 3-35 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Atlanta CSA.
3-103
-------
Boston CSA
60-
40-
-0.1
—I—
0.0
0.1
0.2
0.3
0.4 0.5
Correlation
10
o
a en
Ota 0 85 0/9 0 88 0 79 0 90 0 78 OBI D 78 0 81 0 7
065 085 080 0.90 077 0.90 073 0.60 077 074 0.74
0 08 0 81 0 82 0 80 0 83 0 76 0 85 0 77 0 79 0 72
089 077 0.86 080 062 080 086 065 0.80
-D
c
0
•fi
s
0
O
1.0-
0.9-
0.8
0.7-
0.6-
0.5-
0.4
0.3-
0.2-
0.1-
0.0-
(
K 082 079 ••• B077 0.84 0.82 0 59
090 088 ||J 079 088 ||J 096 084 089 063
'', \f»^- Bo 74 0.8S 078 055
•*^-.J s*Ł"
'*'«.»*••.* 1 I 065
T ** ** % • * • *
.* * * * *•
**. ,:% X
* »4 • . 074 087 0.80 054
* • « 071 II 067
^^H^^^l
* * 07S 059
066
) 50 100 150 200 250 300 350 400 450
Distance (km)
0.77 0.89 061 0.83
07S 088 Obb 083
0 73 IH 0 58 0 86
084 084 075 078
070 076 065 073
066 0.86 060
083 0.83 068 0.77
• ^_
066 IHJ
069 074 066 071
070 084 am 0.82
068 0.81 065 077
0.49
066 0.8S 060
0.54^1
0.49
F
-G
-H
1
-J
-K
•L
-M
-N
-O
-P
-Q
- R
-s
-T
-U
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of the correlations.
Figure 3-36 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max Os in the Boston CSA.
3-104
-------
Los Angeles CSA
150
w
§ 100-
o
0 50
3
1OQ
i7n
1
164
147
148
151
144
87
29
-0.1
0.0
0.1
0.2
0.3
0.4 0.5
Correlation
0.6
0.7
0.8
09
1 0
, <
I
o
...,',«.,
Ti
5«o*|.-r. -^.i-an.f.nfftnna -.« ii' n
1.0
0.9
0.8
0.7
0.6
0.5
0.4-
0.3
0.2
0.1
0.0
-0.1
B "«iB»>!fcl«««'i«m.T..v.., , ««.OD,«i«uiigii
jP^jjjjjjjjj^Jrilj-^^^^" >>g||(i -.
••^^^|)^- :.<• ;."«imfo« ooon^B
50 100 150 200 250 300 350 400 450
Distance (km)
P
I
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of the correlations.
Figure 3-37 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max Q5 in the Los Angeles CSA.
3-105
-------
Atlanta CSA
~c
0
O
30-
25-
20-
15-
10-
5-
33
22
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55
§
fc
>
5
efficient of
0
0
< m
006
0.55-
0.50-
0.45-
0.40-
0.35-
0.30-
0.25-
0.20-
0.15-
0.10-
0.05-
n nn
. ,/Y. *
•\v*. '
'••'
Coefficient of Divergence
OQUJLi-C5I_-5*:
009 009 008 0.09 0.08 0.08 0.11 013 011
010 0.10 008 0,10 009 0.09 0.11 0.13 0.12
010 011 0.11 0.12 011 0.13 0.12 0.13
011 0.07 0.09 010 0.11 010 0.08
011 0,08 005 010 013 012
0.08 010 0.10 0.11 0.07
006 007 012 0.10
010 013 012
013 011
011
-A
-B
-C
-D
- E
-F
-G
-H
-I
- J
-K
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of the CODs.
Figure 3-38 Pair-wise monitor coefficient of divergence (COD) expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max O$ in the
Atlanta CSA.
3-106
-------
Boston CSA
c
13
O
O
100-
80-
60-
40-
20-
O.C
2
75
)0 0.05 0.
114
10 0.
18
15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.5
Coefficient of Divergence
012 014 0.16
§
f
b
§
O
0
0.55-
0.50-
0.45-
0.40-
0.35-
0.30-
0.25-
0.20-
0.15-
0.10-
0.05-
n nn
•
•
•.",
/g?
*f •
.* *'.*
•
i * .
f'V*:
0.06 0.07
007
• •
"t •'/
fl- .
013 0.13 015 014 0.10 017 0.17
010 010 0.07 012 010 006 009
010 0.10 0.07 0.11 010 0.06 009
010 011 007 0.12 008 0.08 007
0.07 0.08 0.08 0,05 0.12 011
0,08 0.08 0,07 012 012
009 0.08 007 0.10
0.09 0.12 0.13
008
0,11
•*
• ^
*
018 012 0.14 0.18 012 019 017 013 019 014
009 012 007 010 007 0.10 011 012 010 011
0.10 0.12 0.08 0.10 0.08 011 011 0.12 011 012
0.09 013 011 010 011 010 0.09 0.13 0.10 014
0.10 0.09 0.12 011 011 015 0.12 010 014 011
011 0.08 0.12 012 011 015 0.13 0.09 015 010
0.09 0.10 0.09 0.09 0.08 0.12 011 0.10 012 010
0,12 0.06 012 012 011 016 013 0.07 016 0.09
0.09 010 010 009 011 012 0.11 010 012 011
0,10 013 0.06 011 007 011 011 013 012 012
0.09 015 0.12 011 013 0.09 0.08 015 0.08 015
014 0.12 0.08 012 012 010 013 012 013
0.12 012 0.10 0.17 014 O.OS 017 0.07
012 0.04 013 0.13 011 013 011
0.11 0.12 0.12 0.12 013 011
013 013 010 014 010
0.11 0.16 0.06 016
014 0.09 015
016 0.06
016
-A
-B
-C
-D
-E
-F
-G
-H
-I
-J
-K
-L
-M
-N
-O
- P
-Q
-R
-S
-T
-u
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of the CODs.
Figure 3-39 Pair-wise monitor coefficient of divergence (COD) expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max Os in the
Boston CSA.
3-107
-------
Los Angeles CSA
400-
„ 300-
o 200-
O
100-
3
155
417
257
181
108
43 16 6 12 16
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45
0.50
0.55
0.00
50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of the CODs.
Figure 3-40 Pair-wise monitor coefficient of divergence (COD) expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max O$ in the
Los Angeles CSA.
3-108
-------
The monitor pairs within the Atlanta CSA (Figure 3-35) were generally well
correlated with correlations between 8-h daily max O3 concentrations ranging from
0.61 to 0.96. The correlations shown in the scatter plot were highest for close
monitor pairs and dropped off with distance in a near-linear form. At a monitor
separation distance of 50 km or less, the correlations ranged from 0.79 to 0.96.
The monitor pairs within the Boston CSA (Figure 3-36) were also generally well
correlated with correlations ranging from 0.49 to 0.96. Again, the correlations shown
in the scatter plot were highest for close monitor pairs, but there was slightly more
scatter in correlation as a function of distance in the Boston CSA compared with the
Atlanta CSA. At a monitor separation distance of 50 km or less, the correlations
ranged from 0.81 to 0.96. The monitor pairs within the Los Angeles CSA
(Figure 3-37) showed a much broader range in correlations, extending from -0.06 to
0.97. At a monitor separation distance of 50 km or less, the correlations shown in the
scatter plot ranged from 0.21 to 0.97. The negative and near-zero correlations were
between monitors with a relatively large separation distance (>150 km), but even
some of the closer monitor pairs were not very highly correlated. For example, Site
AL located at Emma Wood State Beach in Ventura and Site AK situated in an
agricultural valley surrounded by mountains 20 km inland (see map in Figure 3-41)
had a correlation coefficient of only 0.21 over the 2007-2009 warm-season time
period. This was slightly lower than the correlation between Site AL and Site AX on
the Arizona border, 441 km away (R = 0.28). San Francisco and Seattle
(Figure 3-133 and Figure 3-134 in Section 3.9.3) also showed a broad range in pair-
wise correlations, likely resulting from their similar geography where background air
coming in from the Pacific Ocean rapidly mixes with urban pollutants such as NOX
and VOCs from coastal cities and is transported downwind into diversified terrain to
create spatially and temporally varying O3 concentrations.
3-109
-------
s / .r • sui,,i,,,,.: -'
Note: Site AL is near shore, 3 meters above sea level, while Site AK is in an agricultural valley surrounded by mountains,
262 meters above sea level.
Figure 3-41 Terrain map showing the location of two nearby AQS Oz monitoring
sites (red dots) along the western edge of the Los Angeles CSA.
The COD between 8-h daily max O3 measured at paired monitors in all
CSAs/CBSAs (Figure 3-136 through Figure 3-155 in Section 3.9.3) were generally
low, with values similar to those shown in Figure 3-38 and Figure 3-39 for Atlanta
and Boston. This suggests a generally uniform distribution in the 8-h daily max O3
concentration across monitors within these cities and is consistent with the
uniformity observed in the box plots (e.g., Figure 3-32. Figure 3-33. and Figure 3-96
through Figure 3-115 in Section 3.9.2). Los Angeles (Figure 3-34) and San Francisco
(Figure 3-153 in Section 3.9.3). however, had several monitor pairs with COD >0.30
indicating greater spatial heterogeneity. This is consistent with the variability
observed in the box plots for these two CSAs (Figure 3-34 and Figure 3-113 in
Section 3.9.2). In particular, Site AM in the Los Angeles CSA had consistently lower
concentrations (median = 20 ppb, IQR = 17-25 ppb) relative to other sites in the CSA
3-110
-------
(Figure 3-31), resulting in high CODs with other monitors as shown in Figure 3-40.
The O3 monitor at Site AM is located on the Pechanga Tribal Government Building
in Temecula, CA, and began collecting data on June 9, 2008. It is located in a
suburban setting and is classified as a general background monitor. Another close by
site (site ID = 060731201) located in the Pala Reservation, 9.5 km south of this one
(just outside the boundary of the Los Angeles CSA) reported similarly low 2009
8-h daily max O3 concentrations (median = 28 ppb, IQR = 23-32 ppb) between May-
June, 2009 (the only warm-season months with available data from this site between
2007 and 2009).
Note: Site characteristics range from Site A near downtown at 6 meters above sea level to Site D in a forested area on Blue Hill at
192 meters above sea level.
Figure 3-42 Terrain map showing the location of four AQS Os monitoring sites
(red dots) located in or near the city limits in the center of the
Boston CSA.
3-111
-------
There are instances where sites in an urban area may exhibit substantial differences
in median concentrations, but still be moderately well correlated in time. For
example, Sites A and D in Boston (see terrain map in Figure 3-42) have an 11 ppb
difference in median 8-h daily max O3 concentration (COD = 0.16), but a high
correlation (R = 0.90). In this example, Site A is located in the Boston city limits at
an elevation of 6 meters while Site D is located 13 km to the south in a forested area
on Blue Hill, the highest point in Norfolk County (elevation = 192 meters).
The difference in median O3 concentration at these two sites can be attributed to
differing degrees of NOX titration between the neighborhood scale site (Site A) and
the regional scale site (Site D) and to the influence of local topography.
Comparison of monitoring data within the selected focus cities has demonstrated
considerable variability between cities in the behavior of the O3 concentration fields.
Median O3 concentrations vary more within certain urban areas than others.
Likewise, pair-wise monitor statistics (R and COD) are dependent on the urban area
under investigation. These conclusions are consistent with those drawn in the 2006
O3 AQCD (U.S. EPA, 2006b) where a subset of these focus cities were investigated
using similar statistics. As a result, caution should be observed in using data from a
sparse network of ambient O3 monitors to approximate community-scale exposures.
Neighborhood-Scale Variability and the Near-Road Environment
Ozone is a secondary pollutant formed in the atmosphere from precursor emissions
and therefore is generally more regionally homogeneous than primary pollutants
emitted from stationary or mobile point sources. However, O3 titration from primary
NO emissions does result in substantial localized O3 gradients. This is evident in the
near-road environment where fresh NO emissions from motor vehicles titrate O3
present in the urban background air, resulting in an O3 gradient down-wind from the
roadway. Ozone titration occurring in street canyons where NO emissions are
continuously being generated is more efficient because of inhibited transport away
from the source of NO.
Several studies have reported O3 concentrations that increase with increasing
distance from the roadway, both upwind and downwind of the road. Beckerman et al.
(2008) measured O3 profiles in the vicinity of heavily traveled roadways with
Annual Average Daily Traffic (AADT) >340,000 vehicles in Toronto, Canada.
Ozone was observed to increase with increasing distance from the roadway, both
upwind and downwind of the road. This is consistent with scavenging of O3 in the
near-road environment by reaction with NO to form NO2. Upwind of the road,
concentrations were >75% of the maximum observed value at >100 meters from the
road; downwind, concentrations were approximately 60% of the maximum within
200-400 meters of the road. The O3 concentration adjacent to the road on the upwind
side was approximately 40% of the maximum value observed at the site.
Concentrations measured with Ogawa passive samplers over a 1-week period ranged
from 7.3-19.4 ppb with the mean at the two sites ranging from 13.0-14.7 ppb. In a
study of patrol cars during trooper work shifts, Riediker et al. (2003) made
3-112
-------
simultaneous 9-h O3 measurements inside patrol cars, at the roadside, and at a
centrally-located ambient monitoring site. The roadside concentrations were
approximately 81% of the ambient values (mean of 22.8 ppb versus 28.3 ppb). Wind
direction relative to the roadway was not reported.
Johnson (1995) measured O3, NO, and CO concentrations at upwind and downwind
locations near a variety of roadways in Cincinnati, OH. The effects of O3 scavenging
by NO were apparent in the O3 reduction in the interval between 9 meters upwind
and 82 meters downwind of the road. A similar effect was observed by Rodes and
Holland (1981) during an earlier study in which outdoor O3 concentrations were
monitored downwind of a freeway in Los Angeles, CA. In this study, O3
concentrations measured near the roadway were approximately 20% of the
concentrations measured simultaneously at more distant locations judged to be
unaffected by the roadway. Minimal separation distances of the samplers from the
roadway to eliminate measurable influence were estimated to be approximately
400-500 meters for NO, NO2, and O3. Similar results have been observed outside the
U.S., for example in the city of Daegu, Korea, where the yearly roadside
concentrations of CO and SO2 showed a well-defined decreasing trend with distance
from the roadway, whereas concentrations of NO2 and O3 exhibited the reverse trend
(Jo and Park, 2005). During the peak O3 month of May, O3 concentrations in a
residential neighborhood were approximately 40% higher than concentrations at
roadside monitors located 1 meter from the edge of multiple-lane freeways.
3.6.2.2 Rural-Focused Variability and Ground-level Vertical
Gradients
AQS O3 data for monitors located at several rural monitoring sites (e.g., national
parks, national forests, state parks, etc.) were used to investigate rural-focused O3
concentration variability in contrast with the urban-focused variability discussed in
Section 3.6.2.1. These rural monitoring sites tend to be less directly affected by
anthropogenic pollution sources than urban sites. However, they can be regularly
affected by transport of O3 or O3 precursors from upwind urban areas, or by local
anthropogenic sources within the rural areas such as emissions from motor vehicles,
power generation, biomass combustion, or oil and gas operations. As a result,
monitoring data from these rural locations are not unaffected by anthropogenic
emissions.
Six rural focus areas were selected for their geographic distribution across the U.S. as
well as their unique topography and relevance to the ecological assessment in
Chapter 9. Table 3-11 lists the rural focus areas and provides some cursory site
information along with the number of available AQS monitors reporting year-round
and only during the warm-season. Accompanying box plots depicting the distribution
of 2007-2009 warm-season 8-h daily max O3 data from each individual monitor in
the six rural focus areas are included in Figure 3-43. This analysis was restricted to
AQS monitors meeting the same data completeness criteria outlined in Table 3-5 for
3-113
-------
a direct comparison with the 20 urban focus areas investigated in Section 3.6.2.1.
Given the population-center emphasis of the AQS network, limited monitoring sites
(between one and five) were available for each rural focus area. Expanded analyses
of O3 concentrations measured using the more rural-focused CASTNET monitoring
network are included in Chapter 9.
Table 3-11 Rural focus areas.
Focus Area
Adirondack
State Park, NY
Mount Mitchell
State Park, NC
Great Smoky
Mountain
National Park,
NC-TN
Rocky Mountain
National Park,
CO
San Bernardino
National Forest,
CA
Sequoia
National Park,
CA
Short
Name
ADSP
MMSP
SMNP
RMNP
SBNF°
SENP
Year-Round
Os Monitoring
Sites3
1
0
2
1
1
2
Warm-Season
Os Monitoring
Sites'3
0
1
3
0
0
0
Monitor
Elevation
(meters)
1,483
1,982
564 to
2,021
2,743
1,384
560 to
1,890
Site Descriptions
One site on the summit of Whiteface
Mountain in the
Adirondack Mountains
One site near the summit of
Mount Mitchell (highest point in the
eastern U.S.), in the
Appalachian Mountains
Five different locations within
Great Smoky Mountain National
Park in the Appalachia Mountains
One site in a valley at the foot of
Longs Peak in the Rocky Mountains
One site in Lake Gregory Regional
Park (near Crestline, CA) in the
San Bernardino Mountains
Two contrasting sites at different
elevations within Sequoia NP in the
Sierra Nevada Mountains
aNumber of AQS monitors meeting the year-round data set inclusion criteria; the year-round data set is limited to these monitors.
bNumber of AQS monitors meeting the warm-season data set inclusion criteria; the warm-season data set includes May-
September data from both the warm-season and year-round monitors.
°Same AQS site as Site AE in the Los Angeles CSA shown in Figure 3-31.
3-114
-------
Rural Focus Areas
Site ID
360310002
371990004
370870036
470090102
470090101
471550101
471550102
080690007
060710005
061070009
061070006
Years N Mean
07-09 445 50
07-09 447 54
07-09 456 52
07-09 459 47
07-09 459 57
07-09 458 58
07-09 457 60
07-09 456 56
07-09 459 79
07-09 416 76
07-09 459 68
Key
i ns
c. 1o Ł
tn " <=
\—-\ *
SD
13
11
12
12
13
11
11
9
19
16
15
tz
05
1
1
Median
49
54
51
47
57
58
60
56
80
76
69
r*-
|--
IQR Area Site;
16 ADSP A
14 MMSP A
15 SMNP A
16 B
16 C
14 D
13 E
11 RMNP A
28 SBNF A
21 SENP A
19 B
C
to
H
, , I i , , i 1 , i
' "I |r i
"~--\ 4 }•---•
>:..[33-< '
i. - -| i | - - i
h--i 5 ~~H
m,
---
t 1 1 t 1 1
3 50 100 1Ł
03 (ppb)
Note: Includes: Adirondack State Park, NY (ADSP); Mount Mitchell State Park, NC (MMSP); Great Smoky Mountain National Park,
NC-TN (SMNP); Rocky Mountain National Park, Colorado (RMNP); San Bernardino National Forest, CA (SBNF); and Sequoia
National Park, CA (SENP).
Figure 3-43 Rural focus area site information, statistics and box plots for
8-h daily max O3 from AQS monitors meeting the warm-season
data set inclusion criteria within the rural focus areas.
Eastern Rural Focus Areas
In the East, the distribution in warm-season 8-h daily max O3 concentrations from
the Adirondack State Park (ADSP) site on Whiteface Mountain in Upstate NY
(median = 49 ppb) (Figure 3-43) was among the lowest of the rural focus monitors
investigated, but was still higher than concentration distributions measured in the
Boston CSA (medians ranging from 33 to 46 ppb) (Figure 3-33) located 320 km to
the southeast. The ADSP AQS site was included in an analysis for the 2006 O3
AQCD (U.S. EPA. 2006b) and had the lowest year-round median hourly O3
concentration of the rural forested sites investigated (including Yellowstone NP, the
Great Smoky Mountains NP, and Shenandoah NP). For the Appalachian Mountain
monitors in Mount Mitchell State Park, NC (MMSP) and Great Smoky Mountain
National Park, NC-TN (SMNP), there was a fair amount of variability in
concentration distribution. Within SMNP, the median warm-season 8-h daily max O3
concentration ranged from 47 ppb at the lowest elevation site
(elevation = 564 meters; site ID = 470090102) to 60 ppb at the highest elevation site
(elevation = 2,021 meters; site ID = 471550102); these sites are shown on the terrain
map in Figure 3-44. The warm-season median 8-h daily max O3 concentration
gradient between these two sites located 26.2 km apart in SMNP was 0.9 ppb per
100 meters elevation gain.
Data from the five sites within SMNP allowed for further investigation of spatial
variability within the park; Figure 3-45 contains histograms, contour plots and scatter
3-115
-------
plots as a function of distance for the pair-wise correlation and COD (defined in
Equation 3-1) for SMNP. The correlations between the five sites ranged from 0.78 to
0.92 and the CODs ranged from 0.04 to 0.16. The plots of correlation and COD as a
function of distance between SMNP monitor pairs in Figure 3-45 show a large
degree of spatial variability between monitors over relatively short distances. A host
of factors may contribute to these variations, including proximity to local O3
precursor emissions, variations in boundary-layer influences, meteorology and
stratospheric intrusion as a function of elevation, and differences in wind patterns
and transport behavior due to local topography.
Ł§XW "—- U?r
Se»m<"'r ; se««vlle
©
Alcoa ... iF1
Marwtlle - >-
/ v©
V.'ali=ri(l_ _ ...... *
-' V^?' *r;
LI
©
~ S3
' *p "S^^
-,,,:_..-. ^,^©^d
55 S3
Note: The lowest elevation site (Site B) is 564 meters above sea level, while the highest elevation site (Site E) is 2,021 meters
above sea level.
Figure 3-44 Terrain map showing the location of five AQS Os monitoring sites
(green/black stars) in Great Smoky Mountain National Park, NC-TN
(SMNP).
3-116
-------
Smoky Mtn NP, NC-TN
I!
o
-0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Correlation
1 3
O 2
O 1
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55
Coefficient of Divergence
D
1.0
0.9
0.8
0.7
§ °6
;a °5
t 0.4
O 0.3
0.2
0.1
0.0
-0.1
0 50 100 150 200 250 300 350 400 450
Distance (km)
055
0.50
8 o^
c
* 040
§ 0.35-
2 °-3°-
"o
^ 0.25
I "0
I 0,5
O 0.10
0.05
0.00
0.09 0.08 0.08 0.10
0.13
0.15 0.16 B
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histograms includes the number of
monitor pairs per bin and the contour matrix includes the numeric values of the correlations and CODs.
Figure 3-45 Pair-wise monitor correlations (left) and coefficients of divergence
(CODs) (right) expressed as a histogram (top), contour matrix
(middle) and scatter plot vs. distance between monitors (bottom)
for 8-h daily max O$ in Great Smoky Mountain National Park, NC-TN
(SMNP).
Western Rural Focus Areas
The Rocky Mountain National Park (RMNP) site in Colorado at 2,743 meters in
elevation had a warm-season 8-h daily max O3 concentration distribution
(median = 56 ppb, IQR =11 ppb) (Figure 3-43) that is comparable to the
distributions at sites in the Denver CSA located 75 km southeast at elevations around
1,600 meters (medians ranging from 41 to 59 ppb, IQRs ranging from 10 to 16 ppb;
see Figure 3-102 in Section 3.9.1). In nearby Boulder County, Colorado, a 1-year
time-series (Sept 2007-Aug 2008) of ambient surface-level O3 measurements was
collected by Brodin et al. (2010) along an elevation gradient ranging from
1,608 meters to 3,528 meters. The 7 sites used in this study are shown in Figure 3-46
along with the RMNP site and the 15 Denver CSA sites. In fall, winter, and spring,
they observed a clear monotonic increase in O3 concentration with elevation, with a
rate of increase in the mean O3 concentration of 1.5 ppb per 100 meters elevation
gain during winter. In summer, the O3 gradient was similar in magnitude over the
3-117
-------
seven-site transect (1.3 ppb per 100 meters), but much less monotonic; the majority
of the vertical gradient occurred between the lowest two sites (4.5 ppb per
100 meters) and between the highest two sites (5.5 ppb per 100 meters), with the
middle five sites all having approximately equal median O3 concentrations. Ozone
concentrations at the lowest site in Boulder were influenced by NO titration as
evidenced by traffic-related diel cycles in O3 concentrations, but the remaining six
sites were located at elevation in more rural/remote settings and illustrate a positive
surface-level O3 elevation gradient similar to that seen in SMNP and typical of areas
under less direct influence of boundary layer pollution.
©
/g>v S§"©
V:IF'
imp.*,*'-*- ' •;
10km
^City
Bear '
"- K.er carvi
Note: Elevations range from approximately 1,600 meters above sea level in Denver and Boulder, Colorado, to 3,528 meters above
sea level at the highest mountainous site. Blue circles indicate monitoring sites used in the Brodin et al. (2010) study.
Figure 3-46 Terrain map showing the location of the AQS O$ monitoring site in
Rocky Mountain National Park, Colorado (black/green star) and the
Denver, Colorado, CSA (red dots) along with O$ monitoring sites
used in the Brodin et al. (2010) study (blue circles).
3-118
-------
The three sites in California-one in San Bernardino National Forest (SBNF) and two
in Sequoia National Park (SENP)-had the highest distribution of 8-h daily max O3
concentrations of the selected rural focus area monitors included in Figure 3-43.
The SBNF site had a warm-season 8-h daily max O3 concentration mean of 80 ppb
and a maximum of 137 ppb measured on July 1, 2007. This site is located in
Crestline, CA, 90 km down-wind of Los Angeles in the San Bernardino Mountains.
This site was included in the Los Angeles CSA shown in Figure 3-31 (Site AE) and
had the highest median 8-h daily max O3 concentration of any AQS site in the
Los Angeles CSA during this time period (Figure 3-34). This site was also included
in an analysis performed for the 2006 O3 AQCD (U.S. EPA. 2006b) where similarly
high O3 concentrations were observed using 2004 year-round hourly observations.
Note: The lower site (site ID = 061070009) is 560 meters above sea level and the higher site (site ID = 061070006) is 1,890 meters
above sea level.
Figure 3-47 Terrain map showing the location of two AQS O3 monitoring sites
(black/green stars) in Sequoia National Park, CA.
3-119
-------
The two sites in SENP are located 9.7 km apart at contrasting elevations as is
illustrated in the terrain map in Figure 3-47. The correlation in 8-h daily max O3
between these two sites was 0.86 and the COD was 0.09, which are within the range
in correlations and CODs for SMNP (Figure 3-45). The distribution of 8-h daily max
O3 concentrations at the lower elevation site (elevation = 560 meters; site
ID = 061070009) is shifted slightly higher with a median of 76 ppb compared to the
higher elevation site (elevation = 1,890 meters; site ID = 061070006) with a median
of 69 ppb. The lower elevation site is located at the entrance to the park and is at a
low enough elevation to be influenced by boundary layer pollution coming upwind
from Fresno and the San Joaquin Valley. The higher elevation site is in the free
troposphere above the planetary boundary layer and is less influenced by such
pollution. This gives rise to a negative average surface-level elevation gradient of -
0.5 ppb per 100 meters elevation gain in SENP, illustrating the location-specific
complexities inherent to high-altitude surface-level O3 concentrations.
Since O3 produced from emissions in urban areas is transported to more rural
downwind locations, elevated O3 concentrations can occur at considerable distances
from urban centers. In addition, major sources of O3 precursors such as highways,
power plants, biomass combustion, and oil and gas operations are commonly found
in rural areas, adding to the O3 in these areas. Due to lower chemical scavenging in
non-urban areas, O3 tends to persist longer in rural than in urban areas which tends to
lead to higher cumulative exposures in rural areas influenced by anthropogenic
precursor emissions. The persistently high O3 concentrations observed at many of
these rural sites investigated here indicate that cumulative exposures for humans and
vegetation in rural areas can be substantial and often higher than cumulative
exposures in urban areas.
3.6.3 Temporal Variability
3.6.3.1 Multiyear Trends
As reported in the 2010 National Air Quality Status and Trends report (U.S. EPA.
2010e). nation-wide surface-level O3 concentrations in the U.S. have declined
gradually over the last decade. Figure 3-48 shows the downward trend in the annual
4th-highest 8-h daily max O3 concentration from 870 surface-level monitors across
the United States. Figure 3-49 shows a similar trend in the annual 2nd highest
1-h daily max O3 concentration from 875 surface-level monitors. The median annual
4th-highest 8-h daily max dropped from 88 ppb in 1998 to 71 ppb in 2010. Likewise,
the median annual 2nd-highest 1-h daily max dropped from 109 ppb in 1998 to
86 ppb in 2010. The large decreases in 2003 and 2004 in both figures coincides with
NOX emissions reductions resulting from implementation of the NOX State
Implementation Plan (SIP) Call rule, which began in 2003 and was fully
3-120
-------
implemented in 2004. This rule was designed to reduce NOX emissions from power
plants and other large combustion sources in the eastern United States. Reductions in
mobile NOX emissions nationwide from the implementation of recent vehicle and
fuel standards could also be adding to the gradual decline in nationwide surface-level
O3 concentrations (Dallmann and Harley, 2010).
120
100 -
SO-
lo 60 -
o
40 -
20-
98
90 Percentile
• 75m Percentile
50lh Percentile
• 25"1 Percentile
— 10lh Percentile
99
00
01
02
03
04
Year
05
06
07
08
09
10
Figure 3-48 National 8-h daily max Os trend and distribution across 870 U.S.
O3 monitors, 1998-2010 (annual 4th-highest 8-h daily max O3
concentrations in ppm).
3-121
-------
140
120
100-
80-
.3
o
O
8 60-
40-
20-
90 Percentile
. 75lh Percentile
50lh Percentile
. 25lh Percentile
10'h Percentile
98
99
00
01
02
03
04
Year
05
06
07
08
09
10
Figure 3-49 National 1-h daily max O3 trend and distribution across 875 U.S.
Oz monitors, 1998-2010 (annual 2nd-highest 1-h daily max
O3 concentrations in ppm).
The distributional percentiles (10th, 25th, 75th, and 90th) displayed in Figure 3-48
and Figure 3-49 reveal a gradual tightening of the O3 concentration distribution
observed across monitors. For the annual 4th-highest 8-h daily max O3
concentration, the IQR decreased from 17 ppb in 1998 to 9 ppb in 2010. Likewise,
for the annual 2nd highest 1-h daily max O3 concentration, the IQR decreased from
23 ppb in 1998 to 16 ppb in 2010. A similar tightening was observed for the wider
percentiles (90th-10th) for both averaging times.
Weather can have a strong influence on the O3 trends shown in Figure 3-48 and
Figure 3-49. The number of hot, dry days can substantially alter the number of high
O3 days in any given year, even if O3 forming emissions do not change. To better
evaluate the progress and effectiveness of emissions reduction programs, EPA uses a
statistical model to estimate the influence of atypical weather on O3 formation (U.S.
EPA. 2010e). After adjusting for the influence of weather, the downward trend in
surface-level national 8-h daily max O3 concentrations between 2001 and 2008
increased slightly from an 8% reduction prior to adjustment for weather to an 11%
reduction after adjustment for weather (U.S. EPA. 2010e).
A regional breakdown of the trend in O3 concentrations for the 8-hour and 1-hour
metrics is included in Figure 3-50 and Figure 3-51. respectively. In general, the
trends are region-specific with a substantial amount of year-to-year variability.
The reduction in NOX and O3 during the 2003-2004 timeframe is particularly evident
in the North Central and Northeastern U.S. where the NOX SIP Call was focused
3-122
-------
(U.S. EPA, 2010e). The western region (including Alaska and Hawaii but excluding
California) started out with the lowest annual O3 concentration in 1998 and exhibits
the least amount of reduction when compared to 2010 concentrations (11% reduction
in the average annual 4th-highest 8-h daily max and 13% reduction in the average
annual 2nd-highest 1-h daily max). In contrast, California—which has some of the
highest concentrations of the identified regions—shows a larger downward trend in
O3 concentrations over the same time period (19% reduction in the average annual
4th-highest 8-h daily max and 22% reduction in the average annual 2nd-highest
1-h daily max).
100
& 80-
O
70-
60 -
50
Annual fourth highest 8-h daily max
California
West
North Central
Southeast
Northeast
98
99
00
01
02
03
04
Year
05
06
07
08
09
1C
Figure 3-50 Trend in mean 8-h daily max O3 by region, 1998-2010 (mean of the
annual 4th-highest 8-h daily max 63 concentrations in ppm).
3-123
-------
120
Annual second highest 1-h daily max
80 -
70
California
West
North Central
Southeast
Northeast
98
99
00
01
02
03
04
Year
05
06
07
08
09
1C
Figure 3-51 Trend in mean 1-h daily max Os by region, 1998-2010 (mean of the
annual 2nd-highest 1-h daily max Os concentrations in ppm).
Narrowing the focus to changes in O3 concentrations at the individual monitor level,
Figure 3-52 displays the 8-h O3 design value (4th-highest 8-h daily max O3
concentration occurring within a three-year period) for all available monitors for the
2008-2010 period (Figure 3-52A) as well as the change in this design value between
the 2001-2003 period and the 2008-2010 period (Figure 3-52B). Figure 3-53 displays
analogous information for a 1-h O3 design value (4th-highest 1-h daily max O3
concentration occurring within a three-year period). As can be seen in both figures,
the majority of monitors recorded a decrease in design values when comparing the
2001-2003 period to the 2008-2010 period. Specifically, 699 of 853 sites (82%)
included in Figure 3-52B for the 8-h design value and 747 of 869 sites (86%)
included in Figure 3-53B for the 1-h design value reported a decrease of at least
6 ppb in the respective design values. The highest density of monitors reporting
decreases occurs in the Northeast. Only 8 sites (1%) reported an increase of more
than 5 ppb in the 8-h design value and only 16 sites (2%) reported an increase of
more than 5 ppb in the 1-h design value. These sites reporting an increase between
the 2001-2003 and the 2008-2010 periods were located primarily in the West. More
in depth trends analyses have been performed to support the O3 Risk and Exposure
Analysis and are available through EPA's Technology Transfer Network website for
the O3 review (Wells et al.. 2012).
3-124
-------
8-hour Ozone Design Values, 2008-2010
• 40-65 ppb (249 Sites)
• 66-70 ppb (309 Sites)
O 71-75 ppb (303 Sites)
• 76 -90 ppb (168 Sites)
• 91 - 112 ppb (36 Sites)
O
O ' Change in 8-hour Ozone
• | Design Values, 2001-2003 to 2008-2010
O
HAWAII
Increase of 5 to 11 ppb (8 Sites)
Little Change +-5 ppb (146 Sites)
Decrease of 6 to 10 ppb (205 Sites)
Decrease of 11 to 20 ppb (412 Sites)
Decrease of 21 to 31 ppb (82 Sites)
PUERTO RICO
Figure 3-52 Individual monitor 8-h daily max Os design values displayed: (A) for
the 2008-2010 period, and (B) as the change since the 2001-2003
period.
3-125
-------
1-hour Ozone Design Values, 2008-2010
53-80ppb(312 Sites)
• 81-90 ppb (368 Sites)
O 91 - 100 ppb (253 Sites)
O 101 - 120 ppb (140 Sites)
121 - 150 ppb (41 Sites)
Change in 1-hour Ozone
Design Values, 2001-2003 to 2008-2010
Increase of 5 to 14 ppb (16 Sites)
O Little Change+-5 ppb (106 Sites)
O Decrease of 6 to 25 ppb (586 Sites)
O Decrease of 26 to 40 ppb (148 Sites)
Decrease of 41 to 63 ppb (13 Sites)
Figure 3-53 Individual monitor 1-h daily max Os design values displayed: (A) for
the 2008-2010 period, and (B) as the change since the 2001-2003
period.
3-126
-------
Similar findings were reported for regional trends in the 4th-highest 8-h daily max
O3 concentration between 2001 and 2008 in the 2010 National Air Quality Status and
Trends report (U.S. EPA, 2010e). Individual sites that showed the greatest reduction
in O3 between 2001 and 2008 were in or near the following metropolitan areas:
Anderson, IN; Chambersburg, PA; Chicago, IL; Cleveland, OH; Houston, TX;
Michigan City, IN; Milwaukee, WI; New York, NY; Racine, WI; Watertown, NY;
and parts of Los Angeles, CA. Individual sites that showed an increase in O3 over
this time period and had measured concentrations above the O3 standard1 during the
2006-2008 time period were located in or near the following metropolitan areas:
Atlanta, GA; Baton Rouge, LA; Birmingham, AL; Denver, Colorado; El Centra, CA;
San Diego, CA; Seattle, WA; and parts of Los Angeles, CA.
Pegues et al. (2012) investigated changes in 3-year average 8-h daily max O3 design
values between 2003 and 2009 and found reductions at the majority of sites across
the U.S.; consistent with the findings in this section and in the 2010 National Air
Quality Status and Trends report (U.S. EPA, 2010e). Furthermore, they compared
trends in O3 design values between areas that were or were not classified as
nonattainment of the 84 ppb O3 standard in the 2004 designations. Monitors
designated nonattainment achieved O3 design value reductions of 13.3 ppb on
average while monitors designated in attainment achieved reductions of 7.0 ppb on
average.
Looking further back in time, Leibensperger et al. (2008) included an analysis of
June-August 8-h daily max O3 trends from 1980-2006 using AQS data from over
2000 sites in the contiguous United States. They created an index for "pollution
days" representing days when the 8-h daily max O3 concentration was greater than
84 ppb. The observed trend in summertime O3 pollution days over this 27 year
period decreased at an average rate of -0.84 days/year. The authors used several
methods to deconstruct this trend into a component coming from reductions in O3
precursor emissions (-1.50 days/year) and a component coming from climate change
(+0.63 days/year). The climate change impact is a result of decreases in frequency of
mid-latitude cyclones which serve to ventilate surface air over the United States.
Leibensperger et al. (2008) conclude that the reduction in frequency of mid-latitude
cyclones over the 1980-2006 time period has offset almost half of the air quality
gains in the Northeastern U.S. that should have been achieved from reductions of
anthropogenic emissions alone over that period. This conclusion is based on the
assumption of a linear additive relationship between O3 precursor emission changes
and cyclone frequency variations on the rate of change of high O3 days, and does not
account for nonlinearities inherent in O3 chemistry as discussed in Section 3.2.
A more recent analysis by Turner et al. (2012) addressed this issue by utilizing the
GFDL CM3 global coupled chemistry climate model to assess relationships between
summertime cyclones and O3 pollution episodes. They also found a robust decline in
cyclone frequency in modeled scenarios incorporating climate change, but in their
models, less than 10% of the variability in high O3 days in the Northeastern U.S. was
explained by cyclone activity. As a result, Turner et al. (2012) caution against over-
1 The 2008 O3 NAAQS include primary and secondary standards of 0.075 ppm (8-h daily max).
3-127
-------
interpreting the strong association observed by Leibensperger et al. (2008) between
mid-latitude cyclone frequency and the occurrence of high O3 days.
Averaging time can have an impact on perceived trends in surface-level O3
concentrations. Lefohn et al. (2008) investigated the impact of using different
exposure indices on trends in surface-level O3 concentrations in the U.S. by
comparing the annual 2nd-highest 1 -h avg concentration, the annual 4th-highest daily
maximum 8-h avg concentration, and the seasonally corrected 24-h W126
cumulative exposure index. Between 1980 and 2005, most of the urban and rural
sites across the U.S. included in this study showed decreasing or zero trend for all
three of these metrics. However, the magnitude of this trend varied greatly by
exposure index. The largest downward trend in the 1-h and 8-h metrics listed above
were observed in Southern California (>2%/yr downward trend) but the W126
cumulative exposure metric showed large (>2%/yr) downward trends in many
locations across the U.S. including Southern California, the Midwest and Northeast.
By contrasting the 1980 - 2005 trends with more recent 1990 - 2005 trends, Lefohn
et al. (2008) reported that a large number of sites (44%, 35% and 25% of sites for the
1-h, 8-h and W126 metrics, respectively) shifted from a negative trend to no trend.
These shifts in trends were attributed to slow changes in mid-level concentrations
(i.e., 60-90 ppb) following a more rapid change in peak concentrations in the early
years. A similar conclusion was drawn from nationwide O3 data between 1980 -
2008 (Lefohn et al., 201 Ob), suggesting a shift in the O3 distribution over this time
period.
In contrast to the mostly urban observations included in the Pegues et al. (2012)
study above, several studies focusing on rural western monitors have reported
positive trends in O3 concentrations over the last few decades. Jaffe and Ray (2007)
investigated daytime (10 a.m. - 6 p.m. local time) O3 concentrations at rural sites in
the northern and western U.S. between 1987-2004. They found significantly positive
trends in seven of the eleven sites selected ranging from 0.19 ppb/yr in Gothic,
Colorado to 0.51 ppb/yr in Rocky Mountain NP, Colorado (mean trend of
0.26 ppb/yr at these seven sites). No significant trend was observed for the two sites
in Alaska and one site each in Wyoming and Montana. Seasonal analyses were
conducted on the sites having the longest records in Rocky Mountain NP,
Yellowstone, NP and Lassen NP and positive trends were found for all seasons at all
sites. As noted in the 2006 O3 AQCD (U.S. EPA. 2006b). caution should be
exercised in using trends calculated at national parks to infer contributions from
distant sources either inside or outside of North America because of the influence of
regional pollution (see Section 3.4 for a discussion of background O3 concentrations
and international transport).
Trends in baseline O3 concentrations, defined as O3 concentrations at a given site in
the absence of strong local influences, were estimated by region and season in the
U.S. in Chan and Vet (2010). The temperature-adjusted decadal (1997-2006) trends
in estimated baseline O3 varied substantially by region and season. In the Pacific
coastal regions, the trends increased in all seasons except fall, but none of the trends
were statistically significant. In the eastern U.S., negative trends were observed in all
3-128
-------
seasons with the exception of (1) insignificant positive trends in northeast Maine in
summer, fall and winter; (2) significant positive trends in the Midwest in winter; and
3) significant positive trends at one site in Vermont in the summer. The density of
sites in the central and western U.S. were much lower than the coastal and eastern
areas, but in general all sites showed trends that tended to be negative in the spring
and fall but positive in the summer and winter.
Positive trends in marine boundary layer O3 concentrations at several sites on the
Pacific Coast have been reported by other sources in the literature. Parrish et al.
(2009) used observations from multiple coastal sites in California and Washington
and reported a positive annual mean trend of 0.34 ± 0.09 ppb/yr between the mid-
1980s and 2007 (exact dates varied by site depending on available data). A seasonal
stratification of the data at these sites showed the largest positive trend in the spring
(0.46 ±0.13 ppb/yr) with a smaller and non-significant positive trend during fall
(0.12 ± 0.14 ppb/yr). These results agree with positive trends in springtime O3
mixing ratios reported in an earlier study (Jaffe et al., 2003). Positive trends in O3
measurements in the free troposphere above western North America at altitudes of 3-
8 km (above sea level) during April and May of 1995 to 2008 were reported by
Cooper et al. (2010) and discussed in Section 3.4.2 as they relate to intercontinental
transport. Comparable trends were observed in the median as well as 5th, 33rd, 67th,
and 95th percentiles of observations. Note, however, that these results relate to O3
trends above ground level and not to surface O3.
Extending back to the 19th Century, Volz and Kiev (1988) report a series of historic
O3 measurements from Europe. Comparing these with more contemporary
measurements, Parrish et al. (2009) report a 2 to 3 fold increase in boundary layer O3
mixing ratios over the last 130 years with no indication of stabilization in recent
years. Other long-term observations of global trends in the burden of tropospheric O3
as they relate to climate change are discussed in Chapter 10, Section 10.3.3.1.
3.6.3.2 Hourly Variations
Ozone concentrations frequently possess a strong degree of diel variability resulting
from daily patterns in temperature, sunlight, and precursor emissions. Other factors,
such as the relative importance of transport versus local photochemical production
and loss rates, the timing for entrainment of air from the nocturnal residual boundary
layer, and the diurnal variability in mixing layer height also play a role in daily O3
patterns. The 2006 O3 AQCD (U.S. EPA. 2006b) looked at composite urban diel
variations from April to October 2000 to 2004 and found 1-h maxima to occur in
mid-afternoon and 1-h minima to occur in early morning. On a national basis,
however, there was a high degree of spread in these times and caution was raised in
extrapolating results from one city to another in determining the time of day for O3
maxima and minima.
Urban diel variability in O3 concentrations was investigated for the 20 focus cities
listed in Table 3-9 using 1-h avg O3 data from AQS. The year-round data set
3-129
-------
described in Table 3-5 was used to compare diel patterns during cold months
(October - April) and warm months (May - September) between 2007 and 2009.
The warm-season data set, also described in Table 3-5, was used to compare
weekday and weekend diel patterns. Figure 3-156 through Figure 3-160, in the
supplemental material in Section 3.9.4, show these patterns for each of the 20 cities;
examples for Atlanta, Boston, and Los Angeles are shown in Figure 3-54.
Cold Months
Warm Months
Weekdays
Weekends
TO
3
150 -
100 -
50 -
0 -
0 days, 0 year-round sites
mean
median
<=> 5" 95*
<= 1--99"
no year-round data
0 days. 0 year-round sites
no year-round data
327 days. 11 warm-season sites
132 days. 11 warm-season sites
00:00 06:00 12:00 18:00 00.00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
150 -
0
100 -
so -
0 -
637 days, 3 year-round sites
— mean
median
<—=> 5* 95"
459 days. 3 year-round sites
327 days. 21 warm-season sites
132 days 21 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
O
150 -
^ 1 ">o H
en -
c o
< son
0 -
637 days, 47 year-round sites
mean
median
<=> 5*-95*
<=> 1" 991"
459 days. 47 year-round sites
327 days. 50 waim-season sites
132 days, 50 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 00:OC
hour hour hour hour
Note: Uses the year-round data set for the cold month/warm month comparison (left half) and the warm-season data set for the
weekday/weekend comparison (right half). Atlanta had no year-round monitors available for the cold month/warm month
comparison.
Figure 3-54 Diel patterns in 1-h avg O3 for Atlanta, Boston and Los Angeles
between 2007 and 2009.
3-130
-------
In general, all the urban areas showed 1-h daily max concentrations occurring
typically in the early afternoon. In all cities, these afternoon peaks were more
pronounced in the warm months than in the cold months. However, a small peak was
still present during the cold months. During warm months, the difference between the
median daily extrema varied considerably by city. For example, in Los Angeles, the
median 1-h daily min (10 ppb) at -5:00 a.m. was 50 ppb less than the median
1-h daily max (60 ppb) at -2:00 p.m. By contrast, in Boston, the median
1-h daily min (13 ppb) occurred at the same time, but was only 25 ppb less than the
median 1-h daily max (38 ppb). Cities with large daily swings (>40 ppb) in median
1-h O3 concentrations included Atlanta, Birmingham, Los Angeles, Phoenix,
Pittsburgh, and Salt Lake City (Figure 3-156 through Figure 3-160 in Section 3.9.4).
Cities with small daily swings (<25 ppb) in median 1-h O3 concentrations included
Boston, Minneapolis, San Francisco and Seattle (Figure 3-156 through Figure 3-160
in Section 3.9.4). These results are very similar to those found in the 2006 O3 AQCD
(U.S. EPA. 2006b) where many of these same urban areas were investigated. This
supports the conclusions drawn in the previous O3 review that diel patterns in O3
have remained stable over the last 20 years, with times of occurrence of the daily
maxima varying by no more than an hour from year to year.
Using the warm-season data, there was little difference in the median diel profiles for
weekdays compared with weekends across all urban areas. This result stresses the
complexity of O3 formation and the importance of meteorology, entrainment,
biogenic precursor emissions, and transport in addition to anthropogenic precursor
emissions. There was, however, a subtle deviation between weekdays and weekends
in the lower percentiles (1st and 5th) of the distribution. The lower end of the
distribution tended to be lower on weekdays relative to weekends. This is consistent
with analyses in the 2006 O3 AQCD (U.S. EPA. 2006b) and is a result of lower
traffic volumes on weekends relative to weekdays, leading to less NO emissions and
O3 titration on the weekends.
Seasonal and site-to-site variations in diel patterns within a subset of the urban focus
areas presented here were investigated in the 2006 O3 AQCD (U.S. EPA, 2006b).
In northern cities, there was substantial seasonal variability in the diel patterns with
higher extreme values in the O3 distribution during the warm season than during the
cold season. In southern cities, the seasonal differences in extreme O3 concentrations
were much smaller, and some of the highest O3 concentrations in the Houston CSA
were found outside of summer. The general pattern that emerged from investigating
site-to-site variability within the urban areas was that peaks in 1-h avg O3
concentrations are higher and tend to occur later in the day at downwind sites relative
to sites located in the urban core. Differences between sites were not only related to
the distance between them, but also depend on the presence or absence of nearby O3
sources or sinks.
Rural diel variability in O3 concentrations was investigated for the six rural focus
areas listed in Table 3-11 using 1-h avg O3 data from AQS. As with the urban
analysis, the year-round data set described in Table 3-5 was used to compare diel
patterns during cold months (October - April) and warm months (May - September)
3-131
-------
between 2007 and 2009. The warm-season data set, also described in Table 3-5, was
used to compare weekday and weekend diel patterns. Figure 3-55 shows the diel
patterns for each of the rural areas investigated.
There was considerable variability in the diel patterns observed in the six rural focus
areas. The selected mountainous eastern sites in ADSP, MMSP, and SMNP exhibited
a generally flat profile with little hourly variability in the median concentration and
the upper percentiles. In SMNP, there was some diel variability in the lower
percentiles, with higher values during the daylight hours in the warm season data.
This behavior was not present in the data coming from the two year-round monitors
located at lower elevation sites (Sites C and Site D; see map in Figure 3-44),
however, possibly resulting from differing impacts from local sources within SMNP.
For the western rural areas, there was a clear diel pattern to the hourly O3 data with a
peak in concentration in the afternoon similar to those seen in the urban areas in
Figure 3-54 and Figure 3-156 through Figure 3-160 in Section 3.9.4. This was
especially obvious at the SBNF site which sits 90 km east of Los Angeles in the San
Bernardino Mountains at an elevation of 1,384 meters. This site was located here to
monitor O3 transported downwind from major urban areas in the South Coast Air
Basin. It had the highest 2007-2009 median 8-h daily max O3 concentration of any
AQS site in the Los Angeles CSA (see Figure 3-34), and is clearly impacted by the
upwind urban plume which has sufficient time and sunlight to form O3 from
precursor emissions and concentrate the O3 in the shallow boundary layer present at
this elevation.
As with the urban analysis, there was little difference observed in the weekday and
weekend diel profiles using the warm-season data, even down at the lower
percentiles in the distribution. This is consistent with the regional nature of
tropospheric O3. Using the year-round data, there was an upward shift in the
distribution going from the cold months to the warm months, and in some instances
the general shape of the distribution changed considerably as was seen in several
urban sites.
3-132
-------
Cold Months
Warm Months
Weekdays
-s I
"2 <3
I
g
S-
1
00 -
so -
0
414 days, i year-round s«io
.... mean
== 5" -35'"
296 ooys, 1 year-round site
316 days, i
i29 days, i wanri-s*asor* s
00:00 0600 12:OO 18OO OOOOOOOO OB OO 12 OO 1B OO OO OO OO.OO OB 00 12OO 1BOO OOOOOOOO OB OO 12OO 1BOO OO OL
S 8 8
o
0 days 0 yoar- round 5 [os
---- mean
no year-round data
O days O year-round sites
322 Cfayft 1 warm-season sue
128 ddy&, 1 wl
-
cm -season si!e
0000 0800 1200 1800 OOOOOOOO OS OO 12OO 18 OO OOOOOOOO O80O 12OO 180O OOOOOOOO OS OO 12OO 18OO OO OC
Z 150 -
| !«»-
i o
637 days. 2 ysar-round sites
mean
459 days. 2 year-round sites
327 days. 5 warm-season srtes
132 days. 5 warm-season sites
00:00 0600 1300 1800 OOOOOOOO 06 OO UOO 1800 OO'OO OO'OO OS'OO U'OO 1SOO OOOOOOOO 06OO 1100 1800 00 OC
3
l!
150 -
100 -
80
636 days, 1 year-round site
467 M, iy-. 1 ytMi -M ju'iH -.Hi •
326 days. 1 waim-w
131 days, 1 warm-9oasor>8«e
0000 OGCO 1200 1QOO OO OO OO OO OS OO 12OO IS OO OO OO OO:OO OB OO 12 OO 10 OO OO OO OO OO O6 OO 12 OO 15OO 00 OC
632 diyi 1 year-round sit«
459 days. 1 y«ar-round B
327 day* 1 warm-»«a>on *il«
132 days 1 warm-s«ason sit*
0000 060O 120O 18 OO OOOOOOOO 06OO 1200 1800 0000 00 OC WOO 12OO 1BOO OOOOOOOO O6OO 12OO 18 OO OOOC
2 ISO -
z 5 ,00 H
s
629 day*. 2 year-round >it«»
median
5«-95m
• 1--991"
459 day*. 2 year-round »ilea
327 day*. 2 watm-*«a»on »rte*
132 day*, 2 warm-»ea»on tile*
0000 0600 1300 1800 OOOOOOOO 06 OO 13OO 18 OO OOOOOOOO O6 OO 13OO 1SOO OOOOOOOO OS OO 13OO 18OO OOOC
Note: Uses the year-round data set for the cold month/warm month comparison (left half) and the warm-season data set for the
weekday/weekend comparison (right half). Mt. Mitchell SP, NC had no year-round monitors available for the cold month/warm
month comparison.
Figure 3-55 Diel patterns in 1-h avg Os for six rural focus areas between 2007
and 2009.
3-133
-------
3.6.4 Associations with Copollutants
Correlations between O3 and other criteria pollutants are discussed in this section.
Since O3 is a secondary pollutant formed in the atmosphere from precursor
emissions, its correlation with primary pollutants such as CO and NOX can vary
substantially by location. Furthermore, O3 formation is strongly influenced by
meteorology, entrainment, and transport of both O3 and O3 precursors, resulting in a
broad range in correlations with other pollutants which can vary substantially with
season. This section focuses on correlations between O3 and other criteria pollutants
measured at the mostly urban AQS sites: a more detailed discussion of O3 and
O3-precursor relationships is included in Section 3.2.4. To investigate correlations
with copollutants, 8-h daily max O3 from the year-round and warm-season data sets
(Table 3-6 and Table 3-7) were compared with co-located 24-h avg CO, SO2, NO2,
PM2.5 and PMi0 obtained from AQS for 2007-2009. Figure 3-56 and Figure 3-57
contain copollutant box plots of the correlation between co-located monitors for the
year-round data set and the warm-season data set, respectively.
The year-round 8-h daily max O3 data (Figure 3-56) had a very wide range in
correlations with all the 24-h avg copollutants. A clearer pattern emerged when the
data were stratified by season (bottom four plots in Figure 3-56) with mostly
negative correlations in the winter and mostly positive correlations in the summer for
all copollutants. In summer, the IQR in correlations is positive for all copollutants.
However, the median seasonal correlations are still modest at best with the highest
positive correlation at 0.52 for PM2 5 in the summer and the highest negative
correlation at -0.38 for PM25 in the winter. Spring and fall lie in between with spring
having a slightly narrower distribution than fall for all copollutants. The warm-
season 8-h daily max O3 data (Figure 3-57) shows a very similar distribution to the
summer stratification of the year-round data due to their overlap in time periods
(May-Sept and June-Aug, respectively).
3-134
-------
Year-Round
co-
so,-
NO:-
PM2J -I
I
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
Correlation with 8-h Daily Max O,
Winter Spring
CO
S02-
N02 -
H-CED-—H-
--hCD-h-
H—jrrH---
•-H-H^TrH-•••
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0,4 0,6 0.8 1.0 -1,0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0,6 0.8 1.0
Summer hill
-1,0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
Correlation with 8-h Daily Max O,
Note: Year round (Top figure), and with seasonal stratification (Bottom four figures). Shown are the median (red line), mean (green
star), inner-quartile range (box), 5th and 95th percentiles (whiskers) and extremes (black circles).
Figure 3-56 Distribution of Pearson correlation coefficients for comparison of
8-h daily max O3 from the year-round data set with co-located
24-h avg CO, SO2, NO2, PM10 and PM2.5 from AQS, 2007-2009.
3-135
-------
Warm-Season
CO-
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0
Correlation with 8-h Daily Max O3
Note: Shown are the median (red line), mean (green star), inner-quartile range (box), 5th and 95th percentiles (whiskers), and
extremes (black circles).
Figure 3-57 Distribution of Pearson correlation coefficients for comparison of
8-h daily max O$ from the warm-season (May-Sept) data set with
co-located 24-h avg CO, SO2, NO2, PM10 and PM2.s from AQS,
2007-2009.
The seasonal fluctuations in correlations present in Figure 3-56 result in part from the
mixture of primary and secondary sources for the copollutants. For example, O3 is a
secondary pollutant whereas PM2.5 has both primary and secondary origins and these
two pollutants show the largest summertime/wintertime swing in correlation
distributions. This situation arises because the secondary component to PM2.5 is
larger during the summer and is formed in conditions conducive to secondary O3
formation. This results in positive correlations between O3 and PM2.5 during the
summer. During the winter, photochemical production of O3 is much smaller than
during summer and O3 comes mainly from aloft, i.e., the free troposphere (see
Section 3.4.1.1 for further details). In addition, concentrations of PM2.5 are much
lower aloft. On relatively clean days, this can lead to high concentrations of O3 and
lower concentrations of primary pollutants such as PM2 5 or NO. On relatively dirty
days with elevated NO and PM2 5, the intruding O3 is readily titrated by NO in the
boundary layer. These processes result in negative correlations between O3 and
PM2 5 during the winter.
3-136
-------
3.7 Chapter Summary
This section contains a summary of the major topics included in this chapter on the
atmospheric chemistry and ambient concentrations of tropospheric O3 and other
related photochemical oxidants. This chapter has built upon information previously
reported in the 2006 O3 AQCD (U.S. EPA, 2006b) and includes updated material on:
(1) physical and chemical processes of O3 formation and removal; (2) atmospheric
modeling; (3) background O3 concentrations; (4) monitoring techniques and
networks; and (5) ambient concentrations.
3.7.1 Physical and Chemical Processes
Ozone in the troposphere is a secondary pollutant; it is formed by photochemical
reactions of precursor gases and is not directly emitted from specific sources. Ozone
precursor gases originate from both anthropogenic and natural source categories.
Ozone attributed to anthropogenic sources is formed in the atmosphere by
photochemical reactions involving sunlight and precursor pollutants including VOCs,
NOX, and CO. Ozone attributed to natural sources is formed through similar
photochemical reactions involving natural emissions of precursor pollutants from
vegetation, microbes, animals, biomass burning, lightning, and geogenic sources.
The distinction between natural and anthropogenic sources of O3 precursors is often
difficult to make in practice, as human activities affect directly or indirectly
emissions from what would have been considered natural sources during the
pre-industrial era. The formation of O3, other oxidants, and oxidation products from
these precursors is a complex, nonlinear function of many factors including: (1) the
intensity and spectral distribution of sunlight reaching the lower troposphere;
(2) atmospheric mixing; (3) concentrations of precursors in the ambient air and the
rates of chemical reactions of these precursors; and (4) processing on cloud and
aerosol particles.
Ozone is present not only in polluted urban atmospheres but throughout the
troposphere, even in remote areas of the globe. The same basic processes involving
sunlight-driven reactions of NOX, VOCs, and CO contribute to O3 formation
throughout the troposphere. These processes also lead to the formation of other
photochemical products, such as PAN, nitric acid, and sulfuric acid, and to other
compounds, such as formaldehyde and other carbonyl compounds. In urban areas,
NOX, VOCs, and CO are all important for O3 formation. In non-urban vegetated
areas, biogenic VOCs emitted from vegetation tend to be the most important
precursor to O3 formation. In the remote troposphere, methane (structurally the
simplest VOC) and CO are the main carbon-containing precursors to O3 formation.
Ozone is subsequently removed from the troposphere through a number of gas phase
reactions and deposition to surfaces.
Convective processes and turbulence transport O3 and other pollutants both upward
and downward throughout the planetary boundary layer and the free troposphere.
3-137
-------
In many areas of the U.S., O3 and its precursors can be transported over long
distances, aided by vertical mixing. The transport of pollutants downwind of major
urban centers is characterized by the development of urban plumes. Meteorological
conditions, small-scale circulation patterns, localized chemistry, and mountain
barriers can influence mixing on a smaller scale, resulting in frequent heterogeneous
O3 concentrations across individual urban areas.
3.7.2 Atmospheric Modeling
CTMs have been widely used to compute the interactions among atmospheric
pollutants and their transformation products, and the transport and deposition of
pollutants. They have also been widely used to improve basic understanding of
atmospheric chemical processes and to develop control strategies. The domains of
CTMs extend from a few hundred kilometers on a side to the entire globe.
Most major regional (i.e., sub-continental) scale air-related modeling efforts at EPA
rely on the CMAQ modeling system. The horizontal domain for CMAQ typically
extends over North America with efforts underway to extend it over the entire
Northern Hemisphere. The upper boundary for CMAQ is typically set at 100 hPa,
which is located on average at an altitude of ~16 km. CMAQ is most often driven by
the MM5 mesoscale meteorological model, though it may be driven by other
meteorological models including the WRF model and the RAMS. Other major air
quality systems used for regional scale applications include CAMx and WRF/Chem.
Fine scale resolution is necessary to resolve features which can affect pollutant
concentrations such as urban heat island circulation; sea breezes; mountain and
valley breezes; and the nocturnal low-level jet. Horizontal domains are typically
modeled by nesting a finer grid model within a larger domain model of coarser
resolution. Caution must be exercised in using nested models because certain
parameterizations like those for convection might be valid on a relatively coarse grid
scale but may not be valid on finer scales and because incompatibilities can occur at
the model boundaries. The use of finer resolution in CTMs will require advanced
parameterizations of meteorological processes such as boundary layer fluxes, deep
convection, and clouds, and necessitate finer-scale inventories of land use, source
locations, and emission inventories.
Because of the large number of chemical species and reactions that are involved in
the oxidation of realistic mixtures of anthropogenic and biogenic hydrocarbons,
condensed mechanisms must be used to simplify atmospheric models. These
mechanisms can be tested by comparison with smog chamber data. However, the
existing chemical mechanisms often neglect many important processes such as the
formation and subsequent reactions of long-lived carbonyl compounds, the
incorporation of the most recent information on reactions of halogenated species, and
heterogeneous reactions involving cloud droplets and aerosol particles. As a result,
models such as CMAQ have had difficulties with capturing the regional nature of O3
3-138
-------
episodes, in part because of uncertainty in the chemical pathways converting NOX to
isoprene nitrates and recycling of NOX.
Errors in photochemical modeling arise from meteorological, chemical, and
emissions inputs to the model. Algorithms must be used for simulating
meteorological processes that occur on spatial scales smaller than the model's grid
spacing and for calculating the dependence of emissions on meteorology and time.
Large uncertainties exist in the mechanism for oxidizing compounds of importance
for atmospheric chemistry such as isoprene. Appreciable errors in emissions can
occur if inappropriate assumptions are used in these parameterizations.
The performance of CTMs must be evaluated by comparison with field data as part
of a cycle of model evaluations and subsequent improvements. Discrepancies
between model predictions and observations can be used to point out gaps in current
understanding of atmospheric chemistry and to spur improvements in
parameterizations of atmospheric chemical and physical processes.
3.7.3 Background Concentrations
Because the mean tropospheric lifetime of O3 is on the order of a few weeks, O3 can
be transported from continent to continent. The degree of influence from
intercontinental transport varies greatly by location and time. For instance, high
elevation sites are most susceptible to the intercontinental transport of pollution,
particularly during spring. However, because the atmospheric chemistry of O3 is
quite complex and can be highly non-linear in environments close to sources of
precursors, isolating the influence of intercontinental transport of O3 and O3
precursors on urban air quality is particularly problematic.
A number of recent studies indicate that natural sources such as wildfires and
stratospheric intrusions and the intercontinental transport of pollution can
significantly affect O3 air quality in the United States. Two major modeling/field
studies that focused on discerning the contributions of Asian emissions to air quality
in California were the IONS-2010 and the CalNex studies conducted in May through
June of 2010. Modeling and observational components of these studies found
evidence for substantive contributions from stratospheric intrusions and Eurasian
pollution to boundary layer O3. In particular, one modeling study found evidence of
Asian contributions of 8 -15 ppb in surface air during strong transport events in
southern California. These contributions are in addition to contributions from
dominant local pollution sources. Their results suggest that the influence of
background sources on high O3 concentrations at the surface is not always confined
to high elevation sites. It is not clear to what extent the contributions inferred by
these studies are likely to be found in other years, during other seasons, or in other
locations. To gain a broader perspective and to isolate the influence of natural or
transported O3, estimates from CTMs must be used. This is because observations
within the U.S. (even at relatively remote monitoring sites) are impacted by transport
from anthropogenic source regions within the U.S. borders.
3-139
-------
In the context of a review of the NAAQS, it is useful to define background O3
concentrations in a way that distinguishes between concentrations that result from
precursor emissions that are relatively less controllable from those that are relatively
more controllable through U.S. policies. For this assessment, three definitions of
background O3 concentrations are considered, including (1) NA background
(simulated O3 concentrations that would exist in the absence of anthropogenic
emissions from the U.S., Canada and Mexico), (2) U.S. background (simulated O3
concentrations that would exist in the absence of anthropogenic emissions from the
U.S.), and (3) natural background (simulated O3 concentrations in the absence of all
anthropogenic emissions globally). Each definition of background O3 includes
contributions resulting from emissions from natural sources (e.g., stratospheric
intrusion, wildfires, biogenic methane and more short-lived VOC emissions)
throughout the globe. There is no chemical difference between background O3 and
O3 attributable to U.S. or North American anthropogenic sources. However, to
inform policy considerations regarding the current or potential alternative standards,
it is useful to understand how total O3 concentrations can be attributed to different
sources.
Since background O3 concentrations as defined above are a construct that cannot be
directly measured, the range of background O3 concentrations is estimated using
CTMs. For the current assessment, recently published results from Zhang et al.
(2011) using the GEOS-Chem model at 0.5° x 0.667° (-50 km x 50 km) horizontal
resolution and Emery et al. (2012) using a GEOS-Chem/CAMx model (hereafter
referred to as CAMx) at finer horizontal resolution (12 km x 12 km) were used.
Results from these models represent the latest estimates for background O3
concentrations documented in the peer-reviewed literature.
The main results from these modeling efforts can be summarized as follows.
Simulated regional and seasonal means of base-case O3 using both models generally
agree to within a few pbb with observations for most of the United States. However,
neither model is currently capable of simulating day specific base-case O3
concentrations within reasonable bounds. Both models show background
concentrations vary spatially and temporally. NA background concentrations are
generally higher in spring than in summer across the United States. Simulated mean
NA background concentrations are highest in the Intermountain West (i.e., at high
altitude) in spring and in the Southwest in summer. Lowest estimates of
NA background occur in the East in the spring and the Northeast in summer.
NA background concentrations tend to increase with total (i.e., base case) O3
concentrations at high elevation; but that tendency is not as clear at low elevations.
Comparison of NA background and natural background indicate that methane is a
major contributor to NA background O3, accounting for slightly less than half of the
increase in background since the pre-industrial era; and whose relative contribution is
projected to grow in the future. U.S. background concentrations are on average
2.6 ppb higher than NA background concentrations during spring and 2.7 ppb during
summer across the U.S. with highest increases above NA background over the
Northern Tier of New York State (19.1 ppb higher than NA background) in summer.
High values for U.S. background are also found in other areas bordering Canada and
3-140
-------
Mexico. Contributions to background O3 can be episodic or non-episodic; high
background concentrations are driven primarily by the episodic events such as
stratospheric intrusions and wildfires. The most pronounced differences between
these model results and observations are at the upper end of the concentration
distribution, particularly at high elevations. In general, these model simulations
provide a consistent representation of average background concentrations over
seasons and broad spatial areas, but are not able to capture background
concentrations at finer spatial (i.e., urban) and temporal (i.e., specific day) scales.
Note that the calculations of background concentrations presented in this chapter
were formulated to answer the question, "what would O3 concentrations be if there
were no anthropogenic sources." This is different from asking, "how much of the O3
measured or simulated in a given area is due to background contributions." Because
of potentially strong non-linearities—particularly in many urban areas—these
estimates by themselves should not be used to answer the second question posed
above. The extent of these non-linearities will generally depend on location and time,
the strength of concentrated sources, and the nature of the chemical regime. Further
work is needed on how these estimates of background concentrations can be used to
help determine the contributions of background sources of O3 to urban
concentrations.
3.7.4 Monitoring
The FRM for O3 measurement is the CLM and is based on the detection of
chemiluminescence resulting from the reaction of O3 with ethylene gas. Almost all of
the SLAMS that reported data to AQS from 2005 to 2009 used UV absorption
photometer FEMs and greater than 96% of O3 monitors met precision and bias goals
during this period.
State and local monitoring agencies operate O3 monitors at various locations
depending on the area size and typical peak concentrations (expressed in percentages
below, or near the O3 NAAQS). SLAMS make up the ambient air quality monitoring
sites that are primarily needed for NAAQS comparisons and include PAMS, NCore,
and all other State or locally-operated stations except for the monitors designated as
SPMs.
In 2010, there were 1250 SLAMS O3 monitors reporting values to the EPA AQS
database. Since O3 levels decrease appreciably in the colder parts of the year in many
areas, O3 is required to be monitored at SLAMS monitoring sites only during the
"ozone season" which varies by state. PAMS provides more comprehensive data on
O3 in areas classified as serious, severe, or extreme nonattainment for O3. There
were a total of 119 PAMS reporting values to the EPA AQS database in 2009. NCore
is a new multipollutant monitoring network currently being implemented to meet
multiple monitoring objectives. Each state is required to operate at least one NCore
site and the network will consist of about 60 urban and 20 rural sites nationwide.
3-141
-------
CASTNET is a regional monitoring network established to assess trends in acidic
deposition and also provides concentration measurements of O3. CASTNET O3
monitors operate year round and are primarily located in rural areas. At the beginning
of 2010, there were 80 CASTNET sites located in, or near, rural areas. The NFS also
operates a POMS network. The POMS couples the small, low-power O3 monitor
with a data logger, meteorological measurements, and solar power in a self contained
system for monitoring in remote locations. Twenty NFS POMS reported O3 data to
AQS in 2010. A map of the current and proposed rural NCore sites, along with the
CASTNET, and the NFS POMS sites was shown in Figure 3-22.
Satellite observations for O3 are growing as a resource for many purposes, including
model evaluation, assessing emissions reductions, pollutant transport, and air quality
management. Satellite retrievals are conducted using the solar backscatter or thermal
infrared emission spectra and a variety of algorithms. Most satellite measurement
systems have been developed for measurement of the total O3 column. Mathematical
techniques have been developed and must be applied to derive information from
these systems about tropospheric O3. Satellite observations of O3 precursors such as
CO, NO2, and HCHO are also available and useful for constraining model
predictions of emissions of precursors and formation of O3 during intercontinental
transport.
3.7.5 Ambient Concentrations
Ozone is the only photochemical oxidant other than NO2 that is routinely monitored
and for which a comprehensive database exists. Other photochemical oxidants are
typically only measured during special field studies. Therefore, the concentration
analyses contained in this chapter have been limited to widely available O3 data
obtained directly from AQS for the period from 2007 to 2009.
The median 24-h avg, 8-h daily max, and 1-h daily max O3 concentrations across all
U.S. sites reporting data to AQS between 2007 and 2009 were 29, 40, and 44 ppb,
respectively. Representing the upper end of the distribution, the 99th percentiles of
these same metrics across all sites were 60, 80, and 94 ppb, respectively.
To investigate urban-scale O3 variability, 20 focus cities were selected for closer
analysis; these cities were selected based on their importance in O3 epidemiologic
studies and on their geographic distribution across the United States. Several of these
cities had relatively little spatial variability in 8-h daily max O3 concentrations
(e.g., inter-monitor correlations ranging from 0.61 to 0.96 in Atlanta) while other
cities exhibited considerably more variability in O3 concentrations (e.g., inter-
monitor correlations ranging from -0.06 to 0.97 for Los Angeles). The negative and
near-zero correlations in Los Angeles were between monitors with a relatively large
separation distance (>150 km), but even some of the closer monitor pairs were not
very highly correlated. Similar to the correlation, the coefficient of divergence was
found to be highly dependent on the urban area under investigation. As a result,
3-142
-------
caution should be observed in using data from a sparse network of ambient O3
monitors to approximate community-scale exposures.
To investigate rural-focused O3 variability using AQS data, all monitors located
within six rural monitoring areas were examined. These rural monitoring sites are
impacted by transport of O3 or O3 precursors from upwind urban areas, and by local
anthropogenic emissions within the rural areas such as emissions from motor
vehicles, power generation, biomass combustion, or oil and gas operations. As a
result, monitoring data from these rural locations are not unaffected by anthropogenic
emissions. The rural area investigated with the largest number of available AQS
monitors was Great Smoky Mountain National Park in NC and TN where the median
warm-season 8-h daily max O3 concentration ranged from 47 ppb at the lowest
elevation site (elevation = 564 meters; site ID = 470090102) to 60 ppb at the highest
elevation site (elevation = 2,021 meters; site ID = 471550102), with correlations
between the 5 sites ranging from 0.78 to 0.92 and CODs ranging from 0.04 to 0.16.
A host of factors may contribute to variations observed at these rural sites, including
proximity to local O3 precursor emissions, variations in boundary-layer influences,
meteorology and stratospheric intrusion as a function of elevation, and differences in
wind patterns and transport behavior due to local topography.
Since O3 produced from emissions in urban areas is transported to more rural
downwind locations, elevated O3 concentrations can occur at considerable distances
from urban centers. In addition, major sources of O3 precursors such as highways,
power plants, biomass combustion, and oil and gas operations are commonly found
in rural areas, adding to the O3 in these areas. Due to lower chemical scavenging in
non-urban areas, O3 tends to persist longer in rural than in urban areas which tends to
lead to higher cumulative exposures in rural areas influenced by anthropogenic
precursor emissions. The persistently high O3 concentrations observed at many of
these rural sites investigated here indicate that cumulative exposures for humans and
vegetation in rural areas can be substantial and often higher than cumulative
exposures to O3 in urban areas.
Nation-wide surface-level O3 concentrations in the U.S. have declined gradually over
the last decade. A noticeable decrease in O3 concentrations between 2003 and 2004,
particularly in the eastern U.S., coincided with NOX emissions reductions resulting
from implementation of the NOX SIP Call rule, which began in 2003 and was fully
implemented in 2004. This rule was designed to reduce NOX emissions from power
plants and other large combustion sources in the eastern United States. Downward
trends in O3 concentrations in the western U.S. have not been as substantial and
several individual monitors have reported increases in O3 concentrations when 2001-
2003 design values are compared with 2008-2010 design values. Over a longer time
scale, several observational studies investigating O3 concentrations in the marine
layer off the Pacific Coast of the U.S. have reported a steady rise in O3
concentrations over the last few decades. And global scale observations have
indicated a general rise in O3 by a factor of 2 or more since pre-industrial times, as
discussed in Chapter 10. Section 10.3.3.1.
3-143
-------
Urban O3 concentrations show a strong degree of diel variability resulting from daily
patterns in temperature, sunlight, and precursor emissions. Other factors, such as the
relative importance of transport versus local photochemical production and loss rates,
the timing for entrainment of air from the nocturnal residual boundary layer, and the
diurnal variability in mixing layer height also play a role in daily O3 patterns. Urban
diel variations investigated in this assessment show no substantial change in patterns
since the 2006 O3 AQCD (U.S. EPA. 2006b). The 1-h max concentrations tend to
occur in mid-afternoon and 1-h min concentrations tend to occur in early morning,
with more pronounced peaks in the warm months relative to the cold months. There
is city-to-city variability in these times, however, and caution is raised in
extrapolating results from one city to another in determining the time of day for O3
maxima and minima.
Rural O3 concentrations show a varying degree of diel variability depending on their
location relative to larger urban areas. Three rural areas investigated in the east
showed relatively little diel variability, reflecting the regional nature of O3 in the
east. In contrast, three rural areas investigated in the west did display diel variability
resulting from their proximity to fresh urban emissions. These six areas investigated
were selected as illustrative examples and do not represent all rural areas in the U.S.
Since O3 is a secondary pollutant formed in the atmosphere from precursor
emissions, its correlation with primary pollutants such as CO and NOX can vary
substantially by location. Furthermore, O3 formation is strongly influenced by
meteorology, entrainment, and transport of both O3 and O3 precursors, resulting in a
broad range in correlations with other pollutants which can vary substantially with
season. In the copollutant analyses shown in Figure 3-56. the year-round
8-h daily max O3 data exhibited a very wide range in correlations with all the criteria
pollutants. A clearer pattern emerged when the data are stratified by season with
mostly negative correlations in the winter and mostly positive correlations in the
summer for all copollutants. The median seasonal correlations are modest at best
with the highest positive correlation at 0.52 for PM2.5 in the summer and the highest
negative correlation at -0.38 for PM2.5 in the winter. Therefore, statistical analyses
that may be sensitive to correlations between copollutants need to take seasonality
into consideration, particularly when O3 is being investigated.
3.8 Supplemental Information on O3 Model Predictions
This section contains supplemental comparisons between GEOS-Chem simulations
of MDA8 O3 concentrations with observations for 2006 from Zhang et al. (2011) and
Emery et al. (2012). Further details on these simulations can be found in
Section 3.4.3. Figure 3-58 through Figure 3-64 show GEOS-Chem predictions for the
base model (i.e., model including all anthropogenic and natural sources; labeled as
GEOS-Chem in the figure) and the NA background model (i.e., model including
natural sources everywhere in the world and anthropogenic sources outside the U.S.,
Canada, and Mexico; labeled as NA background in the figure) along with
3-144
-------
measurements obtained from selected CASTNET sites (labeled as Measurement in
the figure). Figure 3-65 shows a comparison of GEOS-Chem output with
measurements at Mt. Bachelor, OR, and Trinidad Head, CA, from March-August,
2006. Figure 3-66 shows a comparison of vertical profiles (mean ± 1 standard
deviation) calculated by GEOS-Chem with ozonesondes launched at Trinidad Head,
California and Boulder, Colorado. Figure 3-67 and Figure 3-68 show a comparison
of AM3 simulations of individual stratospheric intrusions during May-June 2010.
Figure 3-69 through Figure 3-74 show box plots for measurements at CASTNET
sites, GEOS-Chem predictions from Zhang et al. (2011) and CAMx predictions from
Emery et al. (2012) for both the base case and NA background. Figure 3-75 shows
time series of AM3 simulations at approximately 2° x 2° at Gothic, Colorado, for
2006.
Connecticut Hill, NY (42N, 76W, 501m)
Acadia NP, ME (44N, 68W, 158m)
100
80
60
40
20
Measurement GEOS-Chem
MA background -
46.7 47.3 25.0
45.3 44.9 24
Huntington Wildlife Forest. NY (43N. 74W, 502m) Kane Exp. Forest, PA (41N, 78W, 622m)
Mar
May
Jun
Jul
Aug
Mar
Apr May
Jun
Jul
Aug
Source: Zhang et al. (2011).
Figure 3-58 Comparison of time series of measurements of MDA8 O3
concentrations at four CASTNET sites in the Northeast with GEOS-
Chem predictions for the base case and for the North American
background case during March-August, 2006.
3-145
-------
Coffeevilie. MS (34N, B9W, 134m)
Sand Mountain, AL (34N, 85W, 352m)
100
80
60
40
20
Measurement GEOS-Chem :
NA background -
56.7 53.3 24.8
Georgia Station, GA (33N, 84W, 270m)
Indian River Lagoon, FL (27N. 80W. 2m)
Mar Apr May Jun Jul
Mar Apr May Jun Jul Aug
Source: Zhang et al. (2011).
Figure 3-59 Comparison of time series of measurements of MDA8 O$
concentrations at four CASTNET sites in the Southeast with GEOS-
Chem predictions for the base case and for the North American
background case during March-August, 2006.
Ann Arbor. Ml (42N. 83W. 267m)
Perkinstown, Wl (45N, 90W. 472m)
Unionville. Ml (43N, 83W, 201m)
Bondville, IL (40N, 88W, 212m)
100
80
60
40
20
49.6 54.8 26.5
Mar Apr May Jun Jul Aug Mar Apr May Jun Jul Aug
Source: Zhang et al. (2011).
Figure 3-60 Comparison of time series of measurements of MDA8 O3
concentrations at four CASTNET sites in the Upper Midwest with
GEOS-Chem predictions for the base case and for the North
American background case during March-August, 2006.
3-146
-------
Yellowstone. WY (45N, 110W. 2400m)
Centennial, WY (41N, 106W, 3178m)
Pinedale, WY (43N, 110W, 2388m)
Rocky Mtn, CO (40N, 106W, 2743m)
I
100
80
GO
40
20
56.8 54.8 40.4
58.1 60.0 42.0
.i......!....., 11.1.,. 11,.,.!.,,,,.... I....i,, 111,
Mar Apr May Jun Jul Aug Mar Apr May Jun Jul Aug
Source: Zhang et al. (2011).
Figure 3-61 Comparison of time series of measurements of MDA8 O$
concentrations at four CASTNET sites in the Intermountain West
with GEOS-Chem predictions for the base case and the North
American background case during March-August, 2006.
Gothic, CO (39N, 107W, 2926m)
Great Basin NP. NIV (39N, 114W, 2060m)
100
80
I *
8 40
20
0
Measurement GEOS-Chem
NA background
56.4 56.1 42.2
56.7 56.3 40.9
Mesa Verde NP, CO (37N, 108W, 2165m) Canyonlands NP, UT (38N. 110W, 1809m)
Mar Apr May Jun Jul Aug Mar Apr May Jun Jul Aug
Source: Zhang et al. (2011).
Figure 3-62 Comparison of time series of measurements of MDA8 O3
concentrations at four CASTNET sites in the Intermountain West
with GEOS-Chem predictions for the base case and the North
American background case during March-August, 2006.
3-147
-------
Grand Canyon NP. AZ (36N, 112W, 2073m)
Big Bend NP, TX (29N, 103W. 1052m)
100
80
60
40
20
Measurement GEOS-Chem
NA background -
58.8 58.9 43 2
Petrified Forest. AZ (35N, 110W, 1723m) Mount Rainier NP, WA (47N. 122W, 415m)
Mar Apr May Jun Jul Aug Mar Apr May Jun Jul Aug
Source: Zhang et al. (2011).
Figure 3-63 Comparison of time series of measurements of MDA8 O$
concentrations at four CASTNET sites in the West with GEOS-
Chem predictions for the base case and the North American
background case during March-August, 2006.
Yosemite NP, CA (38N. 120W, 1605m)
Converse Station, CA (34N, 117W, 1837m)
120
100
8°
60
*
20
0
120
100
8°
60
"o
20
Measurement GEOS-Chem
NA background
64.0 54.0 29.0
68.871.438.1
Death Valley. CA (37N, 117W. 125m)
Trinidad Head, CA (41N, 124W. 107m)
61.257.340.4
41.9 38.3 28.0
Mar Apr May Jun Jul Aug Mar Apr May Jun Jul Aug
Source: Zhang et al. (2011).
Figure 3-64 Comparison of time series of measurements of MDA8 O$
concentrations at three CASTNET sites and the Trinidad Head site
in California with GEOS-Chem predictions for the base case and
the North American background case during March-August, 2006.
3-148
-------
Mt Bachelor (44.0N, 121.7W, 2700m)
Trinidad Head (41.05N, -124.15W, 107m)
observation
GEOS-Chem (0.5x0.667)
GEOS-Chem (2x2.5)
observation (41.9 ppbv)
GEOS-Chem (38.3 ppbv)
Mar
Source: Zhang et al. (2011).
May Jun
Data
Mar
Apr
May
Jun
Jul
Aug
Figure 3-65 Comparison of MDA8 O3 predicted using GEOS-Chem at
0.5° x 0.667° (and 2° x 2.5° resolution; left figure only) with
measurements at Mount Bachelor, Oregon (left); and at Trinidad
Head, California (right) from March to August 2006.
April
August
12
10b
I 8b
03
6 :
4 :
2 :
0
12
10
g 8t
W K k
X) 6
H 4 :
2 :
0
n = 13
n=30
n=31
30 60 90 120 150 30 60 90 120 150
Ozone (ppbv) Ozone (ppbv)
Note: The letter 'n' refers to the number of ozonesonde profiles, and the model was sampled on the same days as the ozonesonde
launches. As can be seen from the figure, variability in both model and measurements increases with altitude, but variability in the
model results is much smaller at high altitudes than seen in the observations at both sites.
Source: Zhang et al. (2011).
Figure 3-66 Comparison of monthly mean (± 1 standard deviation) Os
calculated GEOS-Chem (in red) with ozonesondes (in black) at
Trinidad Head, CA (top) and Boulder, Colorado (bottom) during
April and August 2006.
3-149
-------
RY, 2010-05-28713:32
RH(%)
25 50 75
PS, 2010-05-28714:07
RH(%)
0 25 50 75
SN, 2010-05-28714:41
0 25 50 75
10
8
6
4
2
0
50 100 150 0 50 100 150 0 50 100 150
SH, 2010-05-28713:59 J7, 2010-05-28713:58 J7, 2010-05-29714:00
0 50 100 150 0
O3(ppbv)
50 100 150 0
O3(ppbv)
50 100 150
O3(ppbv)
Note: Shows O3 profiles at multiple sites as observed (black) by ozonesondes and simulated (red) by the GFDL AM3 model at
-50 x 50 km resolution. Also shown are observed relative humidity (gray) and AM3 estimates of O3 concentrations in the absence
of North American anthropogenic emissions (green) and the stratospheric contribution (blue). Model results have been
interpolated to sonde pressure and averaged over 0.5-km altitude bins.
Figure 3-67 A deep stratospheric O$ intrusion over California on May 28 to
May 29, 2010.
3-150
-------
10
8
5
4
2
0
RY, 2010-06-07712:59 PS, 2010-06-07713:56 PS, 2010-06-08713:52 SN, 2010-06-08714:02
RH(%) RH(%) RH(%i RH(%)
0 25 50 75 0 25 50 75 0 25 50 75 0 25 50 75
•q
0 40 80 120 0 40 30 120 0 40 80 120 0 40 80 120
RY, 2010-06-09711:27 PS, 2010-06-09713:59 SH, 2010-06-09714:09 J7, 2010-06-09714:02
0 40 80 120 0 40 30 120 0 40 30 120 0 40 80 120
RY,2010-06-10714:04 PS, 2010-06-10714:21 SH,2010-06-10714:00 J7, 2010-06-10714:00
0 40 30 120 0 40 30 120 0 40 30 120 0 40 30 120
RY, 2010-06-12713:35 PS, 2010-06-12714:00 SH, 2010-06-12714:01 J7, 2010-06-12713:32
0 40 30 120 0 40 30 120 0 40 30 120 0 40 30 120
03(ppbv) 03(ppbv) 03(ppbvJ 03(ppbv)
Note: Shows O3 profiles at multiple sites as observed (black) by ozonesondes and simulated (red) by the GFDL AM3 model at
-50 x 50 km resolution. Also shown are observed relative humidity (gray) and AM3 estimates of O3 concentrations in the absence
of North American anthropogenic emissions (green) and the stratospheric contribution (blue). Model results have been
interpolated to sonde pressure and averaged over 0.5-km altitude bins.
Figure 3-68 A deep stratospheric Os intrusion over California on June 7 to
June 12, 2010.
3-151
-------
Northern half of Eastern sites
GRS420 KEF112
SHN418
CTH110
HWF187
ACA416
Obs
GC
GC-NAB
CX-NAB
Note: Stippled boxes indicate North American background. GRS = Great Smoky NP (North Carolina and Tennessee); KEF = Kane
Exp. Forest (Pennsylvania); SHN = Shenandoah NP (Virginia); CTH = Connecticut Hill Management Wildlife Area (New York);
HWF = Huntington Wildlife Forest (New York); ACA = Acadia NP (Maine).
Source: Adapted from Emery et al. (2012) and Zhang et al. (2011).
Figure 3-69 Box plots showing maximum, interquartile range and minimum
O3 concentrations measured at CASTNET sites (black) in the
Northeast and predictions from GEOS-Chem at ~50 x 50 km
resolution (green) and CAMx at 12 x 12 km resolution (blue) for
May-August 2006.
3-152
-------
Southern half of Eastern sites
SND152 SUM 156 GAS 1 53
IRL141
Obs
GC
'X GC-NAB
cx
* CX-NAB
Note: Stippled boxes indicate North American background. SND = Sand Mountain (Alabama); SUM = Sumatra (Florida);
GAS = Georgia Station (Georgia); IRL = Indian River Lagoon (Florida).
Source: Adapted from Emery et al. (2012) and Zhang et al. (2011)
Figure 3-70 Box plots showing maximum, interquartile range and minimum
Os concentrations measured at CASTNET sites (black) in the
Southeast and predictions from GEOS-Chem at ~50 x 50 km
resolution (green) and CAMx at 12 x 12 km resolution (blue) for
May-August 2006.
3-153
-------
All Central sites
KNZ184
ALC188
CHE185
VOY413
PRK134
CVL151
BVL130
ANA115
UVL124
^
Obs
^
GC
^
cx
***J
GC-NAB
7.".'
CX-NAB
Note: Stippled boxes indicate North American background. KNZ = Konza Prairie (Kansas); ALC = Alabama-Coushatta (Texas);
CHE = Cherokee Nation (Oklahoma); VOY = Voyageurs NP (Minnesota); PRK = Perkinstown (Wisconsin); CVL = Coffeeville
(Mississippi); BVL = Bondsville (Illinois); ANA = Ann Arbor (Michigan); UVL = Unionville (Michigan).
Source: Adapted from Emery et al. (2012) and Zhang et al. (2011).
Figure 3-71 Box plots showing maximum, interquartile range and minimum
Oz concentrations measured at CASTNET sites (black) in the
Central U.S. and predictions from GEOS-Chem at -50 x 50 km
resolution (green) and CAMx at 12 x 12 km resolution (blue) for
May-August 2006.
3-154
-------
Northern half of Rockies sites
GRB411
QLR468
YEL408
PND165
QTH161
CNT169
ROM206
ROM406
Obs
GC
CX
GC-NAB XKi CX-NAB
Note: Stippled boxes indicate North American background. GRB = Great Basin NP (Nevada); GLR = Glacier NP (Montana);
YEL = Yellowstone NP (Wyoming); PND = Pinedale (Wyoming); GTH = Gothic (Colorado); CNT = Centennial (Wyoming);
ROM = Rocky Mountain NP (Colorado, co-located sites).
Source: Adapted from Emery et al. (2012) and Zhang et al. (2011).
Figure 3-72 Box plots showing maximum, interquartile range and minimum
Oz concentrations measured at CASTNET sites (black) in the
Northern Rockies and predictions from GEOS-Chem at ~50 x 50 km
resolution (green) and CAMx at 12 x 12 km resolution (blue) for
May-August 2006.
3-155
-------
Southern half of Rockies sites
o
GRC474
PET427
CAN407
CHA467
MEV405
BBE401
Obs
GC
CX BSS GC-NAB CX-NAB
Note: Stippled boxes indicate North American background. GRC = Grand Canyon NP (Arizona); PET = Petrified Forest (Arizona);
CAN = Canyonlands NP (Utah); CHA = Chiracahua NM (Arizona); MEV = Mesa Verde NP (Colorado); BBE = Big Bend NP
(Texas).
Source: Adapted from Emery et al. (2012) and Zhang et al. (2011).
Figure 3-73 Box plots showing maximum, interquartile range and minimum
Os concentrations measured at CASTNET sites (black) in the
Southern Rockies and predictions from GEOS-Chem at ~50 x 50 km
resolution (green) and CAMx at 12 x 12 km resolution (blue) for
May-August 2006.
3-156
-------
All Western
MOR409
LAV410
YOS404
CON186
DEV412
JOT403
Obs
GC
CX 5S2 GC-NAB SX CX-NAB
Note: Stippled boxes indicate North American background. MOR = Mount Ranier NP (Washington); LAV = Lassen Volcanic NP
(California); YDS = Yosemite NP (California); CON = Converse Station (California); DEV = Death Valley NM (California);
JOT = Joshua Tree NM (California).
Source: Adapted from Emery et al. (2012) and Zhang et al. (2011).
Figure 3-74 Box plots showing maximum, interquartile range and minimum
O3 concentrations measured at CASTNET sites (black) in the West
and predictions from GEOS-Chem at ~50 x 50 km resolution (green)
and CAMx at 12 x 12 km resolution (blue) for May-August 2006.
3-157
-------
Gothic, CO (107W, 39N, 2.9km)
55.9 (7.0)v
58.6 (7.0) r2=0.33 NAbg=42.3 (11.4)
54.9 (4,9) r2=0.06 NAbg=41.1 (6,0)
MAR
APR
MAY
JUN
JUL
AUG
Note: Observed (black) and simulated by the GEOS-Chem (blue; horizontal resolution is 0.5° x 0.667°) and AM3 (red; horizontal
resolution is approximately 2° x 2°) global models. Also shown are the model estimates for North American background (thin
lines); the spike in mid-April likely corresponds to a stratospheric intrusion. The model correlations with observations, average
(over the entire March through August period) total O3 and North American background (NAbg) O3 estimates, and their standard
deviations (shown in parentheses) are presented in the lower left.
Figure 3-75 MDA8 O3 in surface air at Gothic, Colorado for March through
August 2006.
3.9 Supplemental Figures of Observed Ambient O3
Concentrations
3.9.1 Ozone Monitor Maps for the Urban Focus Cities
This section contains supplemental maps showing the location of O3 monitors
reporting to AQS for each of the 20 urban focus cities introduced in Section 3.6.2.1.
The monitors are delineated in the maps as year-round or warm-season based on their
inclusion in the year-round data set and the warm-season data set discussed in
Section 3.6.2.1. The maps also include the CSA/CBSA boundary selected for
monitor inclusion, the location of urban areas and water bodies, the major roadway
network, as well as the population gravity center based on the entire CSA/CBSA and
the individual focus city boundaries. Population gravity center is calculated from the
average longitude and latitude values for the input census tract centroids and
represents the mean center of the population in a given area. Census tract centroids
are weighted by their population during this calculation.
3-158
-------
Legend
Monitor Location*
O VVSrm-season MorilKo
• Yea< round Monilors
• Gty basad Pooulilon Gravity Cantof
• CSA-bawl Population Gravity Cw*V
Urban Ar*m
Atlanta C5A
0 16 30
Figure 3-76 Map of the Atlanta, Georgia, CSA including Os monitor locations,
population gravity centers, urban areas, and major roadways.
Legend
Monitor location.
• Year-found Monitors
• Crty-oased Population Gravity C*nt*r
• CSA-bawO Populaeon Grivtty Cenier
Major HiBhwHy*
BatimoreCSA
Figure 3-77 Map of the Baltimore, Maryland, CSA including Os monitor
locations, population gravity centers, urban areas, and major
roadways.
3-159
-------
Legend
Monitor Location*
O Warm-season Monitors
• Year. lound Monitors
• City-based Peculation Gravity Center
• CSA-basM Population Gravity Cenwr
lB Highways
Figure 3-78 Map of the Birmingham, Alabama, CSA including Os monitor
locations, population gravity centers, urban areas, and major
roadways.
Legend
Monitor Location*
O Worm-season Monitors
• Vaar-rounO Monitor*
• City-baaed Population Gravity Confer
• CSA-based Population Gravity Center
Figure 3-79 Map of the Boston, Massachusetts, CSA including O3 monitor
locations, population gravity centers, urban areas, and major
roadways.
3-160
-------
Legend
Monitor Location*
O VVSrm-seaso
• year-round Monitors
• dry-based Population Gravity
• C SA. b***d Population Gravity
Urban AIMS
Chicago CSA
Figure 3-80 Map of the Chicago, Illinois, CSA including Os monitor locations,
population gravity centers, urban areas, and major roadways.
Legend s
Monitor Location*
O Vtamv&eason Mailtars
• Year-round Mwnlof*
• Crry-baaed Poouiaton Gravfly Center
• CSA-ba*ed PopuUnon Gravity CenUr
Inwntaie Hghways
Ma)w H-ghwaya
|H VMHerBodws
Urban Areas
DiHacCSA
0 20 40
Figure 3-81 Map of the Dallas, Texas, CSA including O2 monitor locations,
population gravity centers, urban areas, and major roadways.
3-161
-------
Legend
Monitor Location*
O Warm-season Monitors
• Year.round Monitors
• C'ly-baMd Population Gravity Comer
• CSA-based Population Gravity Centet
InleislBle Highways
Uw Highways
Figure 3-82 Map of the Denver, Colorado, CSA including Os monitor locations,
population gravity centers, urban areas, and major roadways.
Legend
O ^term-season Monitors
• Ye ai round Monitors
ily-basea Population Crawly O
• CSA-based Populal.cn Gravity C
Interstate Highways
Urban Areas
OetrorlCSA
Figure 3-83 Map of the Detroit, Michigan, CSA including Os monitor locations,
population gravity centers, urban areas, and major roadways.
3-162
-------
Legend
Monitor Location*
O V»rm-sea*on Monitors
• Year-round Monitors
• Cily-baserJ Populalon Gravity C*nt«
• CSA-buad Population Gravity Canter
IntwstHlo Highways
Mapr Highway*
.Wii- !V .'. .
U'Dan Areas
0C8A
Figure 3-84 Map of the Houston, Texas, CSA including Os monitor locations,
population gravity centers, urban areas, and major roadways.
Legend
O Wsim-seasofi Monitors
• Year-round Mention
• CSA-based Population Gravity Center
Figure 3-85 Map of the Los Angeles, California, CSA including Os monitor
locations, population gravity centers, urban areas, and major
roadways.
3-163
-------
Legend
Momlor Location!
O Warm-season Monlort
• Yaw-round Monilcwa
• City-based Population Gravity
• CSA-baaed Population Gravity Cento
interstate Highway*
Mator Highway*
l,Vj;rf Bodms
Urban ATMS
Figure 3-86 Map of the Minneapolis, Minnesota, CSA including Os monitor
locations, population gravity centers, urban areas, and major
roadways.
Legend
Monitor Location*
• Year-round Moniiois
• City-based Population Gravity CBOIBI
• CSA-Oasad Population Gravity Center
kWwsiate Highway*
Major Highways
•A'J-.'T fjOjfi
Urban Areas
0 25 50
Figure 3-87 Map of the New York City, New York, CSA including O3 monitor
locations, population gravity centers, urban areas, and major
roadways.
3-164
-------
Legend
Monitor Loc*Uon»
O Warm-Mason Monitors
• Yoar rouoa Monitors
• City-based Population Gravity Conlitr
• CSA-based Populalion Gravity OrrtW
Interstate Htgfiwavs
Major Highways
H WMM Borfws
Urban Areas
Philadelphia CSA
Figure 3-88 Map of the Philadelphia, Pennsylvania, CSA including Os monitor
locations, population gravity centers, urban areas, and major
roadways.
Legend
Manlier Locillon*
O Warm-season Monitors
• Vsa' round Monitors
• City-based Population Gravity Canter
• CBSA-baMd Population Gravity Cent*
Mafot Highways
Figure 3-89 Map of the Phoenix, Arizona, CBSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways.
3-165
-------
Legend
Monitor Location*
O Wfcrm-saason Monitor*
• Yeai -round Monitors
• Cily-basBd PopulaltDft Gravity Corner
• CSA-bated Populaliw Gravity CsnUr
Interstate Hkgtrwaya
Major Highways
Figure 3-90 Map of the Pittsburgh, Pennsylvania, CSA including Os monitor
locations, population gravity centers, urban areas, and major
roadways.
Legend
Monitor Location*
O Warrn.MBsan Monitors
• Year-round Mon.tors
• City-Cased Population Gravity Cenle<
• CSA based Population Gravity Cantar
inieistate Highways
MOjDCHqhwayt
^H WKar Bodies
Urban ATMS
MLateCSA
Figure 3-91 Map of the Salt Lake City, Utah, CSA including O3 monitor
locations, population gravity centers, urban areas, and major
roadways.
3-166
-------
Legend
Ml I "Mr [ ,„ -,r,i n .
O Warm-seasa
• Year-round Monitors
• City-bawd Population Gravity Cento'
• CBS A based Population Gravity Cen
IntefUata Highways
Msjot Highway*
San Antonio CBSA
Figure 3-92 Map of the San Antonio, Texas, CBSA including Os monitor
locations, population gravity centers, urban areas, and major
roadways.
Legend
Monitor Location*
O Warm-season Monitors
• Year-round Monitors
• City-MMd Population Gravity Center
M Population Gravity Cenier
Urban ATM*
San Franosco CSA
Figure 3-93 Map of the San Francisco, California, CSA including O3 monitor
locations, population gravity centers, urban areas, and major
roadways.
3-167
-------
Legend
O WBrnvsesson Monrlois
• Year-round Monitors
• City-based Population Gravity
• CSA-baaM Population Gravity Ci
krtwstaW Highways
Major Highway*
.V.HiH' Bodies
Urt;jn Aieas
SeatttoCSA
25 80
Figure 3-94 Map of the Seattle, Washington, CSA including Os monitor
locations, population gravity centers, urban areas, and major
roadways.
Legend
Monitor Location*
O Warm-season Monito's
• VMt-round Ntofiiiois
• City-based Population Gravity Centm
• CSA-bawd Population Gravity Canlw
lftlef*tate Highyrmyj
Mapsi Highways
iV.m-i h-. •-•;
UiOan Areas
Si Lout CSA
Figure 3-95 Map of the St. Louis, Missouri, CSA including O3 monitor locations,
population gravity centers, urban areas, and major roadways.
3-168
-------
3.9.2 Ozone Concentration Box Plots for the Urban Focus Cities
This section contains box plots depicting the distribution of 2007-2009 warm-season
8-h daily max O3 data from each individual monitor in the 20 urban focus cities
introduced in Section 3.6.2.1. Monitor information including the AQS site id, the
years containing qualifying data between 2007 and 2009, and the number of
8-h daily max O3 observations included in the data set are listed next to the box plot.
Statistics including the mean, standard deviation (SD), median and inner quartile
range (IQR) are also shown for each monitor with the site letter corresponding to the
sites listed in the figures above.
Site ID
131210055
130890002
131350002
130670003
132470001
131130001
131510002
130770002
130850001
132230003
Years
07-09
07-09
07-09
07-09
07-08
07-09
07-09
07-09
07-09
Key
^ —
N
450
452
446
450
455
306
459
455
458
455
C
JC 00
Ł« I
cs E
-\ •
Mean
53
52
52
51
51
52
51
47
47
50
median
SD
18
16
16
18
15
15
17
16
13
14
1
ojc
i
Atlanta CSA
Median IQR Site , , , , i , , , , i , , , ,
54 22
52 23
52 18
52 22
51 22
52 20
51 22
47 19
47 17
50 21
*LO
r~-
I"-
-H
A-
B-
c-
D-
E-
F-
G —
H-
I-
J-
K-
C
; -I * I ^
• ^ i- — i — J ^ •
if I 1
.' f .'
( — % — I _ "*
• i jl i
.' f H
" r_ OJ] H '
-A
-B
-C
-D
-E
-F
^™ f^
-H
-I
- J
- K
) 50 100 150
03 (ppb)
Figure 3-96 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Atlanta, Georgia, CSA.
3-169
-------
Baltimore CSA
Site ID
245100054
240053001
240051007
240330030
240251001
240030014
240130001
240313001
110010025
110010041
110010043
240259001
240338003
^1 ni ^nnoA
O I U I JUUiiU
510595001
515100009
510591005
240210037
510590030
510590018
511071005
24009001 1
510590005
240170010
511530009
511790001
510690010
510610002
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-09
08-09
07-09
07-09
07-09
07-09
07-09
AT— no
u/ uy
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
Ł
N
454
456
459
445
450
459
459
292
453
459
459
458
452
459
456
432
458
459
459
456
439
459
456
453
459
459
456
C
Ij E
\ •
Mean
42
51
46
50
54
53
50
49
49
50
51
52
51
C-l
O I
50
47
52
51
52
51
52
51
49
51
49
48
46
44
c
— Ljrr}— -<
H-HZD •
~L 1
-1 1 !
:- 1 \4 I <
!- 1 )
j. — | ;
5- 1
1 f~~
"••••"LZ
,l [^
1
1
} i
\ i
(- i
| — - - ^
rj. i
^--- -H
1
:- 1 fr }- 1
t [ fi }-- - - 1
'!---- - 1 if 1 '.
"r • f | <
' i_J_J---!
1 [__|i_ J- ---\
r----L~4 I H
:- \ ^ \ - - -4
;- 1 i [--M
;- - - - -( ty ^ ^
;- - - - -| 4 j- - - - i
:- - - - ) j| 1 *
; 1 4; 1 -i
!•---[ 4 :)---•!
-A
-B
-C
-D
- E
- F
-G
-H
- I
- J
-K
"™ L
-M
Kf
IN
-0
-p
-Q
-R
- S
._ ~T
-U
-V
-w
-X
-Y
-Z
-AA
-AB
0 50 100 150
03 (ppb)
Figure 3-97 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Baltimore, Maryland, CSA.
Birmingham CSA
Site ID
010730023
010731003
010736002
010732006
011170004
010731010
010731005
010735002
010735003
010731009
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
&
t - -
N
450
459
459
455
459
452
459
459
456
457
c
j: a:
to g
I *
Mean
47
45
48
48
49
47
47
47
48
46
c
to
XJ
---! f h---H
h--r~fi-----i
-A
-B
-c
-D
™ C
__ p
___ /"N
-H
- I
- J
1 ' ' ' i ' ' ' • i • ' ' '
) 50 100 150
03 (ppb)
Figure 3-98 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Birmingham, Alabama, CSA.
3-170
-------
Boston CSA
Site ID
250250042
250250041
250092006
250213003
250171102
250170009
250095005
330111011
250270024
250094004
440071010
250270015
3301 10020
330150016
330115001
330150014
250051002
440030002
330131007
440090007
330012004
Years
07-09
08-09
07-09
07-09
07-09
07-09
07-09
07-09
09
07-08
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
^> K
> -
N
459
306
459
459
457
439
459
457
153
305
453
458
455
458
459
459
459
458
459
459
459
c
re
E
•
Mean
35
42
44
46
41
41
44
40
38
46
46
47
38
41
46
40
46
44
39
46
39
C
E
I
SD
13
12
15
15
15
14
14
13
12
14
15
15
12
13
13
12
13
15
12
14
11
if
>**
Median IQR Site , , .
33
41
41
44
40
39
42
37
38
43
44
46
36
40
44
39
45
42
37
45
37
R
I „ .
16
17
20
20
20
19
18
18
16
18
21
19
16
16
17
16
19
18
16
19
15
"g
~ 4
A-
B-
c-
D
E-
p _
G-
H-
l-
j-
K-
L-
M-
N_
0-
P-
Q-
p =.
s-
T-
U-
I- • -Q]
v..
:--- p
r - -[]
>-c
h " C
• --
>..
:- -
;--
}- • Ł
-,.,
• ~ ~c
!>..
0
1 ... > I ....
l> *j- •«-«.;
jjjj 1 „ .{
-j?-J i
==f=\- H
If I -:
-pFi <
3D""1
L3*_J- 4
r^jr-i .<
1 J i> J- i
"pTfl •?
Ji_ J- ----- -i
[^jjj™^} J-
^JF] •;
LLJLJ---^
I—HE—I j
i jr I •••••J-
I : fr. i i
-A
r- B
-C
D
-E
— C
i- Q
-H
hi
-J
- K
h- L
- M
k- N
^ 0
^P
-Q
H- R
,_ g
,_T
i- U
50 100 150
Of _ _ L \
3 (ppb)
Figure 3-99 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Boston, Massachusetts, CSA.
Chicago CSA
Site ID
170314002
170311003
170310076
170310042
170310072
170310064
170436001
170310001
170314007
170311601
170310032
170317002
170314201
180890030
180892008
1 70890005
180890022
171110001
181270024
170971002
170971007
550590019
181270026
171971011
180910005
180910010
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
&
i- • • -
N
458
452
458
412
459
459
459
459
458
457
450
450
611
440
451
459
455
458
456
459
459
457
453
458
453
456
c <®
S i
•\ •
Mean
39
44
44
45
42
41
39
46
39
48
46
43
42
45
45
44
42
43
46
41
46
47
43
42
42
44
'•5
i
I
SD
13
13
14
14
12
13
12
14
13
14
13
13
13
15
13
13
13
12
14
13
13
14
13
12
12
13
P
oie
Median IQR Site
38 18 A-
43 17 B-
44 18 C-
44 17 D -
42 17 E-
40 18 F-
39 16 G -
46 19 H -
37 18 I -
47 19 J -
45 17 K -
42 17 L -
41 17 M-
44 19 N-
44 18 0 -
42 16 P -
41 15 Q-
42 15 R-
44 17 S -
39 18 T -
46 18 U-
45 19 V-
42 18 W-
41 15 X-
41 15 Y-
43 17 Z-
1 .,.,!,,,.
I- • • p~|tT-| ... |
:- - - -j ^ | i
§.._<
j
-J.
--t
-<
^
:----! |»: }•--•<
:- - -- - j - ^ ~1 i
t---CDD---<
r - — | [T 1 ^
t---| > ( 1
:- 1 T^ | ^
» — dm---*
:-----! h 1 <
:•---! j I----H
•?.,, ) L j*» , ». ^
?---OEI3 — <
S----LJ! | -:
1- - - - ^ . ^ \ - - - -:
t-'-QEUr '
> - - - j h ] <
• ' " -1 fr 1- - - -i
f - - -QLJ- - • <
>---rf~\---i
- A
*- B
-C
i- D
— Ł
i- F
-G
^ H
- 1
- J
- K
-L
^ M
^ N
r-Q
r P
-Q
™ R
r- S
i- T
-U
r- V
i- W
>- x
- Y
-Z
, .... I ....
0 50 100 150
>- o>
I ~ - ~ - 1
03 (ppb)
Figure 3-100 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Chicago, Illinois, CSA.
3-171
-------
Dallas CSA
Site ID
481130069
481130075
481130087
484393009
484393011
484392003
480850005
481390016
484391002
483970001
481210034
484390075
482570005
481211032
482510003
481391044
482311006
483670081
482210001
Years
07-08
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
08-09
07-09
07-09
07-09
Key
!•---
N
279
456
429
459
457
455
456
455
458
449
456
459
459
459
459
306
459
459
459
c
f, K
%> §
csi Ł
Mean
41
48
47
48
46
52
52
43
46
47
52
52
47
50
47
47
43
48
44
.1
1
SD
14
15
16
16
15
16
14
14
16
13
15
16
12
13
15
12
12
14
15
If
§;|
Median IQR
38 22
46 21
44 24
46 23
44 22
50 23
51 21
42 22
44 23
47 21
50 22
50 24
45 18
49 19
45 22
45 19
42 18
47 22
41 22
1b
-c
in
Site
A-
B-
C-
D~
E~
F -
G-
H-
I-
J -
K-
L-
M-
N -
0-
P-
Q-
B =
s-
c
1 1 J .,,,,,,. ,1..,,,,..
;---[ [•_
j..,Ł^j
f - -| J
!- - - Ł^
±- - 1 ]i
J-- - { ]
i- - . r~~™™
"-n
:----t
;- -- j
5- - -|
--Lfc
-FT?
)
H • J i }— -< 1
i i 1 ( 1 I 1 1 t 1
fc---<
p>-<
&T"1 .
Pr';'
7f—\- — <
» [• -:
• I - - - ^
D— -i
FD....;
-A
-B
-C
~D
^E
- F
-G
-H
-I
- J
-K
-L
-M
-N
-0
-P
-Q
-R
~S
i i i i i i i i i i
50 100 150
03 (ppb)
Figure 3-101 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Dallas, Texas, CSA.
Denver CSA
Site ID
080310025
080310002
080310014
080013001
080590002
080590005
080050002
080590011
080350004
080590006
080590013
080130011
080137001
080137002
081230009
Years
08-09
07
07-09
07-09
07-09
07-09
07,09
07-09
07-09
07-09
09
07-09
07
07
07-09
Key
T__.
N
299
153
450
441
459
456
306
459
456
457
150
453
152
142
451
To 3
8 E
-\ •
Mean
49
39
51
55
54
55
54
56
58
60
50
56
42
56
55
.1
*
E
i
SD
12
10
12
11
12
12
11
12
11
12
9
12
10
11
11
!l
oE
Median IQR Site
51
41
52
57
56
56
55
57
58
59
50
56
42
56
56
c.
r-
(""
16
13
15
13
16
16
14
15
14
15
10
14
12
13
14
0)
A-
B-
c-
D-
C ™
F-
G-
H-
I-
J-
K-
L-
M-
N-
0_
| , , , , I
'--— |_sLJ— '
^---{^f}--i
'. { ^ ',
i Q
i. - ~ _ ^ | ^
• 1 j
» — • -| ^
;- 1 j
H — n
j. — rj
h Łj
^--m--!
f — i_i
i i i i i
0 50
H*"1
D"^
^]- H
|- « ^
ZZF " ~:
| f — -:
^^--^
^ [-- i
ZD — J|
_|---i
^}"i
-A
-B
-C
-D
-E
-F
-G
-H
- |
-J
- K
-L
-M
-N
_0
100 150
03 (ppb)
Figure 3-102 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Denver, Colorado, CSA.
3-172
-------
Detroit CSA
Site ID
261250001
261630019
260991003
261630001
261610008
260990009
260490021
261470005
260492001
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
^r>
(.--.
N
459
456
452
459
459
459
458
459
455
c
Ł ra
CM E
-f •
Mean
46
47
47
42
45
46
44
43
45
C
vs
|_
SD
14
15
15
13
13
15
13
15
14
1
1
*
Median IQR Site
46 18
46 19
46 18
41 16
44 17
45 18
44 19
41 19
45 18
r-
| —
?
A-
p ™
C-
D-
C _
p _
G-
H -
I-
C
1 1 C 1
J"----] H
f. — J T
rS
:-~S
~ — i ^
)
i i , , , i i i , i
> i- i.
i [• ^
1- ~,
50 100 M
03 (ppb)
-A
, p
-C
-D
-E
-F
-G
-I
?0
Figure 3-103 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Detroit, Michigan, CSA.
Houston CSA
Site ID
482010075
482010070
482010066
482010047
482010055
482010416
482010046
482011035
482010051
482010024
482011034
482010062
480391004
482010026
482011039
482011015
482010029
482011050
483390078
481671034
480391016
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
N
430
451
433
451
441
451
454
437
445
455
443
450
430
454
442
428
449
444
455
427
455
mean
*
Mean
36
34
37
38
40
38
37
35
38
45
39
34
38
41
41
38
45
40
43
37
34
median
1
SD
17
16
17
16
18
17
16
17
17
17
16
16
18
16
18
15
16
17
12
17
16
lif
ol
Median IQR Site
32 25
30 24
32 25
35 22
35 25
34 26
34 23
31 24
33 25
43 24
37 22
28 24
33 26
39 23
36 27
33 21
42 22
35 27
42 16
33 27
28 23
-«>
[..
..4
A-
p _
c-
D-
C »_
F_
G-
H-
t
i ~~
J ™
K-
L-
M-
N-
o-
P-
Q-
R-
S —
T-
U-
C
^ T J * f i
' j t? ' _ '
' -TT« — I
"" "j ''i? '. H
^' — I;* . ' — , 4
, ' i .u i — *
'^:;;::r
i- -[™]T*~~~^ -t
^_iSzocLOircoH-D
i i i i i i i i i i i i i i i i i i i i i
) 50 100 150
03 (ppb)
Figure 3-104 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Houston, Texas, CSA.
3-173
-------
Los Angeles CSA
Site ID
060371602
060371301
060371302
060371103
060372005
060374002
060595001
060590007
060375005
060371002
060370002
060370113
060370016
060371701
060591003
060371201
060711004
060376012
060650004
060592022
061112002
060658005
060712002
060658001
061110007
060710012
060379033
061110009
060719004
060659001
060710005
060656001
060714003
060714001
060710306
061113001
061111004
061112003
060650009
060650012
060651016
060710001
060655001
060719002
060652002
060651999
060651010
060711234
060650008
060659003
Years
07-09
07-08
09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
08
07-09
07-09
08-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
09
07-09
07-09
07-09
07-09
07-09
07-09
07-08
09
07-09
07,09
07-09
Key
Ł
(....
N
458
306
152
457
459
459
459
459
459
459
459
459
458
459
459
459
457
457
127
457
455
276
459
440
459
456
452
458
457
453
459
459
459
455
459
453
458
457
153
457
459
455
459
452
448
283
153
453
265
444
:H?Sb « '
1 ( J4_ L — L, ^
' •>- \ u i
' \ fc i ~"
,, ,J L 1 j
• .1 c 1 1
'j . * , I ^
' • ' — I A 1 ^
I -r-t- ' "*
'»«•'
"LLi ^ — —
^-rrS=l"""'
^$^'
-A
-B
-C
-D
-E
-F
-G
-H
-I
""* J
-K
-L
-M
-N
-0
™ p
-R
-S
-T
-U
- V
-w
-X
- Y
-z
-AA
-AB
-AC
-AD
-AE
-AF
-AG
-AH
-Al
-AJ
-AK
-AL
-AM
-AN
-AO
- AP
-AQ
-AR
-AS
-AT
-AU
-AV
-AW
-AX
) 50 100 150
03 (ppb)
Figure 3-105 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Los Angeles, California, CSA.
3-174
-------
Minneapolis CSA
Site ID Years N Mean SD Median IQR Site
270031002 07-09
271390505 07-09
271636015 07-09
271713201 07-09
270031001 07-09
551091002 07-09
270495302 07-09
271453052 07-09
Key
*"?...
456
459
439
446
455
457
454
453
.c re
to »
CM E
"\ *
41
42
43
42
39
43
44
39
c
ro
E
I
12
11
12
11
12
11
10
11
P
lit
41
42
42
43
38
42
44
39
to
| „ „
16
14
16
16
18
15
14
15
"tft
iite
A-
B -
C-
D-
E -
P _
G-
H-
I , I I , f , J 1 , , , r
:- - - j f |- - - 1
t- | f | <
t - - - |_J_} - - H
>• i i~i <
"• • " -I Ą } - - '
; -- {_J_J- - - <
[, . . . | ^ [ - . 4
•r - - -f ^f]-- ^
' ' ' ' I ' ' ' ' 1 ' ' ' '
-A
-B
-C
-D
_ Ł
- F
_ Q
- H
0 50 100 150
Oj(ppb)
Figure 3-106 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Minneapolis, Minnesota, CSA.
New York CSA
Site ID
360810124
360610135
360050110
360050133
340030006
3401 70006
340130003
360850067
361192004
090010017
361030002
340315001
340250005
34023001 1
340273001
090019003
361030009
340190001
360790005
340210005
090013007
090011123
340290006
360715001
361030004
090090027
360270007
090093002
090050005
361111005
Years
07-09
08-09
07-09
07-09
08-09
07-09
09
07.09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
s
«o
> - . -
N
446
298
457
459
300
442
122
298
444
447
454
445
458
459
456
457
890
455
459
456
457
459
456
459
453
456
456
459
446
459
c
j- CO
fn I
CN E
-\ •
Mean
43
39
40
41
42
45
36
45
46
49
47
45
47
48
48
47
47
50
44
49
49
46
51
45
48
41
43
47
46
41
c
.™
-o
I
SD
15
15
15
14
15
17
14
16
17
15
15
15
15
17
16
16
16
16
14
16
15
17
16
14
14
14
14
15
15
12
_*:
"SS
is
o;E
!
Median IQR Site
41 21
38 19
39 19
39 18
41 20
43 20
36 19
43 23
44 22
47 20
46 20
43 19
45 19
47 22
47 22
44 21
46 20
48 21
42 19
48 22
47 19
43 22
49 20
43 17
46 18
40 17
41 17
45 18
43 19
39 16
*«
h~
!•-
^
en
- ^
A-
B-
C -
D-
E-
F -
G-
H-
I -
J-
K-
L-
M-
N -
0-
P-
Q-
R-
S-
T ™
u-
V-
w-
X-
Y-
z-
AA-
AB-
AC^
AD-
C
• ' •_Lj^LJJ, i i i i 1 i i • i
^--QLJ -*
1 ^1
=--^p_----:
1S|
>r-^Q»»»'*
>.-r~w~\""^
^ ^
t . -i i» t <
^--CED — <
-A
- B
-C
- D
- E
- F
-G
- H
- I
-J
- K
- L
-M
_ N
-0
- P
-Q
-R
-S
-T
- U
- V
- w
-X
- Y
-z
-AA
-AB
-AC
-AD
i i , i i . i i i j . i i .
) 50 100 150
03 (ppb)
Figure 3-107 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the New York City, New York, CSA.
3-175
-------
Philadelphia CSA
Site ID
421010014
421010004
340070003
420910013
340150002
421010024
420450002
420170012
100031013
100031010
340071001
420290100
3*110007
100031007
420110006
4201 1001 1
Years
07
07-09
07-08
07-09
07-09
07-09
07-09
08-09
07-09
07-08
07-09
07-09
07-09
07-09
08-09
08-09
Key
to ^
N
153
459
298
454
433
429
458
305
455
304
450
457
458
450
306
306
c
i E
Mean
50
39
51
51
50
51
49
48
48
53
52
50
50
49
44
47
§
e
SD
15
13
17
16
16
16
15
16
15
16
15
15
1ft
lo
14
15
13
14
fl
O'f"
Median 1QR Site
48
38
51
50
50
49
49
47
48
51
52
50
Ł.1
O I
50
48
43
46
t^
22
17
23
22
21
21
20
22
20
21
20
20
1 O
I 3
21
20
17
19
Ł
^ {^__*^_l ; I - - - - *
A ^
B-
C-
D-
E-
F -
G-
H-
I -
J-
K -
L -
M ~
N-
O -
P -
Q -
i , , , 1 i , , a 1 , i , ,
..y^j^ "
*• i > ^ <
:- - - - {]^;f^l -i
i Q Infill ----- ^
>•--- -LJfe_J H
;" " • -[ jjt } -'
> - - - { 'f'Z~l " " * ' "*
J. - . .. (,,^1^;;; - - 'i
i---^Q^3 i
••"•""" [IltlLlTI ^
^•CD--<
L... r=y=|^_. '
^ • ' - • 1 ^ 1 • ' • - <
t---Lj:J-----i
-A
-B
-C
- 0
-E
- F
- G
r- H
- 1
~ J
-K
- L
-N
- O
- P
- Q
0 50 100 150
Of rv_-U\
3 (PPD)
Figure 3-108 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Philadelphia, Pennsylvania, CSA.
Phoenix CBSA
Site ID
040133002
040133003
040139997
040131004
040134005
040134003
040130019
040137020
040137024
040137022
040132001
040137021
040134004
040131010
040137003
040132005
040139704
040135100
040134010
040134008
040139702
040139706
040213001
040213010
040213009
040217001
040139508
040134011
040213003
040218001
040213007
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
Ł
h - --
N
455
459
455
459
454
459
459
459
459
457
459
457
459
456
455
459
459
453
457
459
451
448
459
458
459
459
459
459
459
459
459
i- G
s 1
H •
Mean
53
57
56
58
55
55
55
56
56
56
53
59
56
55
52
57
58
55
48
58
53
58
59
45
48
52
57
46
52
59
50
^6
1
1
SD
9
10
10
10
10
9
10
9
9
10
10
9
9
9
8
8
9
10
9
9
9
11
9
9
9
9
8
9
9
9
8
aS
^
5.<
:-{3D--<
j -- -MLJ--- <
'• - • ffVn - ~ •'
i — rf~i--<
^ - - j t j- - -:
•' - - ryi - - <
:---f--ryn--<
>-EU-<
i--{33--<
.--op-.
>-{jO--^
-^'ffl,-'1
!...^.1
:- — OD-i
^-CC-H
>--CE-i
>-OD-i
*--{3H-<
t-.Q3.M
>--m;-<
^•CE--^
•,^3-<
^{TB-M
i--CEr-<
^-OJ--'
,...133..,
f--cp-i
1 . 1 1 '-t-* 1 1 1 1 1 1 f • t 1
-A
-B
-C
-D
-E
-F
-G
^H
^ 1
- J
-K
^L
-M
-N
^0
i-P
-Q
^R
>-S
-T
-U
i- V
-w
-X
i- Y
-Z
i- AA
^-AB
-AC
r- AD
-AE
0 50 100 150
O3 (ppb)
Figure 3-109 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Phoenix, Arizona, CBSA.
3-176
-------
Pittsburgh CSA
Site ID
420030008
420030010
420030067
421290006
420031005
421250005
421255001
421250200
421290008
420070005
420070014
420070002
420050001
420730015
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
N
459
459
449
459
450
459
449
459
456
452
459
452
459
459
c
!M E
-) •
Mean
49
48
49
45
50
48
48
46
47
47
46
49
50
45
median
I
SD
14
13
12
13
15
12
13
12
13
13
13
13
15
13
1
]
Median IQR Site
48 19
47 18
49 16
44 18
50 20
48 17
47 17
46 16
46 17
47 16
46 18
49 16
48 21
44 19
Ł«
}--
-H
zSi— TCt — iO"nrnoocD>
„ i i i i i i i i i i i i i i
, , , i i , , , « i , , , i
r---\ f 1 ^
^--Pfl-.-M
>---! h 1 -i
J 2 — I j
i 1j r •
I— -|— fT-)--M
) 50 100 1Ł
03 (ppb)
-H
A-
B-
c-
D-
E-
F-
G-
H-
1-
J-
K-
L-
C
, l.
I
i--f » l--"i
V/3B
V-S
__4S
> 50
[ 1
H""1
}--1
J--H
-A
-B
-C
-D
-E
-F
-G
-H
-I
-J
-L
100 150
03 (ppb)
Figure 3-111 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Salt Lake City, Utah, CSA.
3-177
-------
San Antonio CBSA
Site ID
480290055
480290032
480290052
480290622
480290059
Years
08-09
07-09
07-09
08-09
07-09
Key
1-
N
306
454
456
305
450
1 1
- { •
Mean
40
42
43
37
36
'•5
1
SD
14
15
13
13
13
1
Median IQR Site
37 20
39 20
41 18
33 20
33 18
fo
\ - '
en
-1
m O o oo >
_ 1 1 1 I 1
, , , , I , , , , 1 , , , ,
!• ,^
i--[2
)
|i j <
;[• 1 1
™^~j» \ -i
"f ] -;
• — i *
< CD O Q UJ
1 « 1 1 I
50 100 150
03 (ppb)
Figure 3-112 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the San Antonio, Texas, CBSA.
San Francisco CSA
Site ID
060010009
060750005
060010006
060012004
060012001
060811001
060131004
060011001
060130002
060410001
060950006
060010007
060852007
060950004
060133001
060850005
060851001
060950005
060131002
060550003
060870006
060870003
060953003
060870007
060970003
060852006
060870004
060850002
060971003
060690002
060690003
Years
08-09
07-09
07-08
08-09
07-09
07-09
07-08
07-09
07-09
07-09
07-08
07-09
07-08
07-09
07-08
07-09
07-09
07-09
07-09
07-09
07-08
07-09
07-09
07-09
07-09
07-09
07-09
07-09
08-09
07-09
07-09
Key
j, , „ „
N
306
458
303
306
459
459
306
456
458
458
306
459
306
459
306
456
459
459
459
459
306
456
455
456
459
458
459
459
306
456
457
C
^ ra
"Ł E
r\ *
Mean
29
28
31
25
35
31
31
34
42
29
40
43
34
35
41
36
39
39
47
37
38
31
44
33
31
44
33
44
34
42
54
«3
•3
ai
E
I
SD
9
8
9
7
10
9
8
10
13
8
11
14
10
10
10
10
12
11
12
9
9
8
13
8
8
11
8
11
10
10
12
l|
o|E
Median IQR Site , , . , i , , , , i , , , ,
28
27
30
24
33
29
29
33
40
28
39
41
33
34
41
35
37
37
45
35
37
30
43
32
31
43
32
42
33
40
54
F~
\- "
12
10
12
10
12
11
12
12
18
10
13
18
13
12
12
13
16
12
15
10
11
10
17
10
10
14
10
15
13
12
16
..?
A-
B-
/"* _
D-
p — ,
C _
/"* ._
H-
I-
J-
K-
E __
M-
N-
O-
P-
Q-
R-
S-
T-
u -
V -
w-
X-
Y-
z
AA-
AB-
AC~
AD-
AE-
!•-[_Ł];--<
;• - 1 ^ i ;- - 1
>-QTJ--i
i -m- { j
f.-,j
•-*.
-H
,..<
i- - • 1 ; [» t — ^
\ - ijfj,- - J:
i - - 1; k 1 — t
1 - - [i )» | H
'"CO"""1
:- - | fc \ ---',
'. - f f~\ 1
>"-Cf3'"J-
^-Tijii I -:
"•"DE ^
;. - . [~li~} ^
I---I j» |----<
:--| '[» 1 H
v - -fF"! - - 1
:•--! ^ ] <
h - l_fc} - - f
;--ŁJJ--J.
'r ' J.l E 1 •
'"i.KJ"""
:~THE3 — 1
M-|_J|_J"-H
:---[ [» I- '.
y--- \ | ', ---i
i , , 1 , Wpfe^jJ , j f . 1 j j 1
-A
- B
_ |*>
-D
-E
— P
— f""
- H
- I
- J
- K
-L
-M
- N
-O
- P
-Q
— R
-s
-T
- U
- V
- w
-X
- Y
-2
- AA
- AB
-AC
-AD
- AE
0 50 100 150
Os (ppb)
Figure 3-113 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the San Francisco, California, CSA.
3-178
-------
Seattle CSA
Site ID
530330080
530330010
530330017
530330023
530670005
530531008
530531010
530530012
530570018
530570013
Years
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07,09
07-08
07
Key
Ł
U"}
N
452
456
432
444
459
443
459
286
279
153
c
c. Ifl
13 I
-\ •
Mean
28
32
36
38
35
35
31
39
26
30
median
1
SD
8
12
12
14
10
12
11
9
8
10
A-
B-
c-
D-
E-
F-
G-
H -
I-
J-
C
, , I , , , , I
1P~
;- - - ] |i | i
_A
-c
-D
-E
-F
-G
-H
-I
-J
> i 1 i | i iii | i i i i
) 50 100 150
03 {ppb)
Figure 3-114 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the Seattle, Washington, CSA.
St. Louis CSA
Site ID
295100086
295100085
291890004
291890014
171630010
290990019
290990012
291831002
291831004
171193007
171190008
171191009
291890005
170831001
291130003
171170002
Years
07-08
07-09
07-08
07-09
07-09
08-09
07
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
07-09
Key
Ł
N
302
459
459
765
444
306
153
449
459
458
452
458
755
459
457
457
*« |
oj E
-\ *
Mean
50
46
50
48
43
49
55
51
50
48
48
50
46
46
50
46
median
SD
16
14
15
13
13
12
16
14
13
14
13
14
12
12
13
11
His
i
Median IQR Site
50 19
46 18
51 18
48 16
44 17
49 16
53 19
49 18
49 15
48 17
48 17
49 18
46 16
45 16
49 15
46 14
Ł
(--
-1
A-
D .^,,
c-
D-
E-
G-
H-
I
i
J-
K-
L-
M™
N-
o-
P-
C
l ,,<<[, i ,
V-.fi
y~~;^fi
:-'--&
'- — j '
f. — |
!• 1 4
'---US.
^j — 4
^ \ — i
i] — i
• \---i
i i — '.
^j ^
i f -i
^j - - - H
^j .;
> [ i
-A
K»w R
-C
-D
-E
-F
-H
-1
- J
-K
-L
-M
-0
-P
) 50 100 150
O3 (ppb)
Figure 3-115 Site information, statistics and box plots for 8-h daily max O3 from
AQS monitors meeting the warm-season data set inclusion criteria
within the St. Louis, Missouri, CSA.
3-179
-------
3.9.3 Ozone Concentration Relationships for the Urban Focus Cities
This section contains histograms and contour matrices of the Pearson correlation
coefficient (R) and the coefficient of divergence (COD) between 8-h daily max O3
concentrations from each monitor pair within the 20 urban focus cities discussed in
Section 3.6.2.1. These figures also contain scatter plots of R and COD as a function
of straight-line distance between monitor pairs.
20
c 15-
O ID-
S'
-0.1 0.0 0.1 0.2 0.3 0.4 05 0.6 0.7 0.8 0.9 1.0
Correlation
1.0-
0.9-
08-
0.7-
0.6-
0.5-
0.4-
0.3-
0.2-
0 1 •
00-
-0.1
0 75 0 76
090 087 074 075
0.79 077 073 075 078 0.79
090 082 077 081 081 038
085 068 0.71
061 076
063 070
50 100 150 200 250 300 350 400 450
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-116 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max Os in the Atlanta, Georgia,
CSA.
3-180
-------
Baltimore CSA
I
o
200-
150-
100-
50-
-c
1 0
2 9
0 0.1 0.2 0.3 0.4 0.5 06 0
Correlation
88
7 0
209
8 0
70
1.0
AB
50 100 150 200 250 300 350 400 450
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-117 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max Os in the Baltimore, Maryland,
CSA.
3-181
-------
Birmingham CSA
25-
„ 20-
§ 15'
O 10-
5-
-0.1 0.0 0.1 0.2 0.3
0.4 05
Correlation
0.6 0.7 0.8 09 1.0
CD O
1.0-
0.9-
0.8-
0.7-
0.6
0.5-
0.4-
0.3-
0.2-
0.1
o.o-I
-0.1
-c
-D
-E
F
G
-H
J
50 100 150 200 250 300 350 400 450
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-118 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max Os in the Birmingham,
Alabama, CSA.
3-182
-------
Boston CSA
_
g 40-
°20-
-0.1
00
0 1
02
03
0.4 0.5
Correlation
06
0 7
08
0.9
< m o o in
O x _
ZOG.OCEOTI-D
1.0-
09-
08-
0.7-
06-
05 -
JD
&
8 0.4
03-
0.2-
0.1 -
0.0-
-0.1
090 085 0.84 0.88 077 0.83 0.88 079 0.82 076 0,84 076 086 069
096 085 MROBO 084 BH 083 089 079 088 079 090 078
082 084 «|jj 085 08S 060 090 077 080 073
083 OB9 087 OJB OB8 081 082 080 083 076
082 094 :;•-.• . OM089 077 0.88 0 BO OS2
082 OK :.!'_, OS) 0.92 0/v 0 8J 082 a t-'j
0% 090 079 0.88 ^H 0.88 0.84 0.89 063
0.88 0 78 0 75 0.88 ^H 0.74 0.85 0.78 0 55
0.87 ^H 0.88 0.78 0.85 0.78 0 73
0.79 0.77 ^1 0.71 ^1 0 65
084 071 073 0.73 07? 081
0.8C 076 0.88 079 004
U.B/
„
074 076
081 078
080 077
085 077
030 086
0 77 0.89
075 0.88
0.73
0.84 0,84
070 076
050
083 083
069 0.7'
061 07.
074 0
079 0
065
o ei
065
058
0.75 0.78
065 0.'
060 060
./o
»
068 077
0.85 060
50 100 150 200 250 300
Distance (km)
350 400 450
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-119 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max Os in the Boston,
Massachusetts, CSA.
3-183
-------
Chicago CSA
„ 150-
1 100-
o
50-
-0
1.0 -
0.9-
0.8-
0.7-
06-
c
•2 05-
1
a>
t
O n A -
O u>t
0.3-
0.2-
0.1 -
0.0-
n 1
U. I
(
199
69
— _ — 52
5 I !•
1 00 0.1 0.2 0.3 0.4 0.5 0.6 07 0.8 0.9
Correlation
09J09S07708609008eMo8«087087083086080(Blio870830870e7080085080080077083083
|| 0 : 0 91 094 0 91 0 92 0 94 0 S3 Ml' 0 87 0 83 0 at 0 91 0 90 0 88 0 90 Ml 0 88 0.88 0.86 0.84 0 79 0.84 0 87
0 78 0 88 0.94 0 91 0 95 0 91 0.01 0 89 0.86 0.80 O.SsBBo.SS 0.88 0.88^0 86 0.87 0 85 0.88 0.81 0.85 0.89
0 SI 0 79 0 76 0 75 0.80 0 76 0 81 0 80 0 70 0 74 0 76 0.80 0 74 0 79 0 76 0 69 0 78 0 73 0 76 0.72 0 77 0 73
M40B50B6090083^B^OBOOB60BBOa508BOB50890870880870B1 a 77083085
0 6ei|).91 0 92 0 57 0 95 0.89 0 80 0.9ollo.86^lo,B5^^Ho.89 0.88 0.95 0.78 0.84 0.88
0 90^B^Boe4 061 0 74 084 0 OsHJD 84 ^Bo 66 084 082 081 0.85 087 0 77 0.84
0 87 Mil 068083078086 OS1 0 88 0 89 0 87 0 89 0 83 0 83 0 81 0 86 0 8 1 0.84 0 67
^^* ^^H
V. . 089 0.88 087 076 0.87 086 069 086 088 036 ••0.88 OS8 0 S3 079 077 0.85
«lf • •
'yftftl >'. > 0.840.81 074 0830.85^8 0.85 0.86 0.87 0.83 0.82 0.81 0.840.850780.84
'•""gjfilij^."^ , _
• XV*4.»%'»t:.5MJ'">. 076085085081083081084085088080080072081081
L* ••.' •••<"/." "•
?~r ,' • . 0.63 0.82 0.80 073 0,82 0 75 0 71 0.8S 0 77 0 65 0.66 0.80 0 73
*' . *"••"'
• * OJ8^Bo.85^Bo.84 0.87 0.84 0.88 077 0.88 0^8
OS:'0»0840SI 084081 081 084079061
0 80^0 82 0 81 0 80 0 87 0 79 0.84 0 68
0 83 0 81 0.88 0 83 0 77 0 81 0.80 0 80
0 84 0 83 0.83 jljo 79 0.88 ^J
I^IHo 80071 071 083
^|o 79 072 080 082
0.7B0.71 0.750.82
079 0.82 JH
0 70 0.79
H
) 50 100 150 200 250 300 350 400 450
Distance (km)
|
1.0
B
D
E
F
G
H
I
J
K
-L
-M
N
•O
p
Q
p
U
V
w
X
Y
z
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-120 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max Os in the Chicago, Illinois, CSA.
3-184
-------
Dallas CSA
c
D
O
80-
60-
40-
20-
-c
1
1.1 0.0 0.1 0.2 0.3 0.4 0.5 0
Correlation
11
6 0
34
7 0
92
8 0
CQ O Q LJJ
2 z o Q. a cc
g
75
2
0
O
1
0
0
0
0
0
0
0
0
0
0.89 0.90 090 0.91 0.90 085 089 0.91 0.90 081 083 0.85
090 0.92 086 0.92 093 084 0.89 0.93 089 0.89 0.85
0-
9-
8-
7-
6^
5-
4-
3-
2 -
1 -
0.0-
-n
-i
I^^M^^^^^^^^^^H
086 0.93 088 080 0.93 0.90 0.89 079 0.86 0.87
k^^HHHHH^^^HHB •••••1
0.96 0.90 084 0.94 089 0.93 0.91 080
0.90 076 0.95 0.95 0.85 0.79 0.85 0.83
088 085 0.96 0.88 0.92 0.97 0.83
•1H
L*«y 0.73 0.82 0.90 IB • 0.85 0.80
.. AiCfcA.^
• •"Wi ' 089 084 075 0.81 0.87
' -^Vv'V."
*.*..••• f ••.**. . 0.85 086 0.93 082
. *• '
• .
» .« « 0.85 0.85 0.89
•
"
0.81
•
0.70 0.80 0.67
0.84 0.77 0,81
0.71 0.87 0.89
0.83 0.78 0.79
069 0.90 0.88
0.81 0.83 0.80
• 0.67 0.70
^^^^^
0 65 091 0 93
0.73 0 87 0.85
0.82 0.76 0.80
0.88 071 0.73
0.79 0.62 0.78
0.73 0.80 0.89
0.61 0.63
•
0.86
0.88
0.83
0.87
0.81
0.85
0.87
0.79
0.83
0.93
0.84
0.83
0.88
0.82
0.73
0.79
0.82
0.80
0.83
0.83
0.85
0.89
0.72
0.83
0.90
0.78
079
0.92
0.79
0.66
0.88
0.80
0.77
0.82
0.69
:
0.77
:
0.61
0.88
„,,
0.72
067
0.76
0.77
0.55
•
0.71
-A
-B
-C
-E
-F
-G
-H
_ I
-J
-K
•L
-M
-N
-0
-P
-Q
-R
-S
50 100 150 200 250 300 350 400 450
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-121 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max Q5 in the Dallas, Texas, CSA.
3-185
-------
Denver CSA
50-
_ 40-
§ 30-
5 20-
10
10
-0.1 0.0 0.1 0.2 03 0.4 0.5 0.6 0.7 08 09 10
Correlation
-------
Detroit CSA
20-
E 15-
§10-
5-
-0.1 0.0 0,1 0.2 0.3
0.4 0.5
Correlation
0.6 0.7 0.8 0.9 1.0
o
1.0
0.9-
0.8-
07
0.6
I 0.5-
I
0.3-
0.2
0.1 •
00
-0.1
••'.*
I - A
0.64 - B
c
-D
-E
-F
-G
-H
50 100 150 200 250 300 350 400 450
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-123 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O$ in the Detroit, Michigan,
CSA.
3-187
-------
Houston CSA
§
80-
60-
40-
~c
2 1°
1 00 0.1 02 03 04 05 0
Correlation
6 0
33
7 0
88
8 0.
< co O O
u. O I _
1.0-
0.9
0.8
0.7
0.6
05-
04-
0.3
0.2-
0.1 -
0.0-
-0.1
059 094 098 Q.78
. 082 0.89 HI 0.87 0.88 HI 0.87 0.79 0,38 0.08 0.76
089 081 I Hlogi 074 HI OK 088 083 080 069 065
: HHIoas o.eaHloe9 o.at 084 os? O.TS
^^•83 ^S B 0 85 0 78 087 DM 0 80
0.76 068 090 082 084 086 0.76 0.77 0.57
0 us 0 90 0 83 0 80 0 6S 0.64
0 83 093 0 86 0 76 089 : JC OK!
077 088 078 0 73 088 OSI 0«2
Q.B4 0.69 0.66
061 0.81
077 DM 063 aea
073 077 0&5
057 0.87
0.46
A
•B
C
•D
E
F
084 076 ^H 088 I
^1 0.83 ^H !
077 0.88 076 0
090 094 C
• °
K
-L
M
N
0
P
50
100 150
200 250 300
Distance (km)
350 400 450
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-124 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O3 in the Houston, Texas, CSA.
3-188
-------
Los Angeles CSA
150 -
§ 100-
o
0 50-
3
63
109
17D
164
147
148
151
144
87
-0.1
0.0
0.1
0.2
0.3
0.4 0.5
Correlation
0.6
0.7
29
0.8
0.3-
0.2-
0.1 -
0.0-
-0.1
••--vx-
/ . • •••••/. . ; .
.•.*':•, •.••••
• •. --.A. v:
• ••.•••
0 50 100 150 200 250 300 350 400 450
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-125 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max Q5 in the Los Angeles,
California, CSA.
3-189
-------
Minneapolis CSA
15-
o
O
5-
18
-0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6
Correlation
0.7 0.8 0.9 1.0
O
LU
O
1.0-
0.9-
0.8-
0.7-
0.6-
I 0.5-
I
0)
Q 0.4 H
0.3-
0.2-
0.1 -
o.o-
-0.1
-G
0 50 100 150 200 250 300 350 400 450
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-126 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O$ in the Minneapolis,
Minnesota, CSA.
3-190
-------
New York CSA
c
i
o
150-
100-
50 -
-C
1 6 | 20
.1 0.0 0.1 0.2 0.3 0.4 0.5 0
Correlation
Dt
6 0
7 0.8
0.9
< co o Q
03 O Q
-0.1
0 50 100 150 200 250 300 350 400 450
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-127 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max Os in the New York City, New
York, CSA.
3-191
-------
Philadelphia CSA
~ 60"
I 40-
°20-
-0.1
00
0 1
02
03
0.4 0.5
Correlation
06
O 0. O
1.0-
09-
08-
0.7-
06-
0.5 -
JD
&
8 0.4
03-
0.2-
0.1 -
0.0-
-0.1
H 091 087
•••• 096 094 093
077 0.86 086 077 062
082 068 086 062 065 082 OS
083 068 086 085 089 084 06
:195 093 095 091 080 00! 091 088 081 084 089
0.68 079 081 082 060 081 062 076 06
0.90 0-89 082 088 088 081 083 087 05
097 092 38-1 093 091 0 85 090 086 06
0.87 084 090 089 082 083 0.87
flr. 0 91 091 0 fl-1 0 9t 0 83 OB
76 089 088 080 067 0.80 0.8
0.78 0.86 0.82 069 0.7
078 0,65 0,90
-
086 085 086
067 071
::
50 100 150 200 250 300 350 400 450
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-128 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O$ in the Philadelphia,
Pennsylvania, CSA.
3-192
-------
Phoenix CBSA
g 100-
3
o
0 50-
-c
30
1 | r~
.1 0.0 0.1 0.2 0.3 0.4 0.5 0
Correlation
125
6 0
14"i
7 0
122
27
_B^^H
8 0.9 1.0
< CD O D UJ LL.
50
100 150
200 250 300
Distance (km)
350 400 450
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-129 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max Os in the Phoenix, Arizona,
CBSA.
3-193
-------
Pittsburgh CSA
40-
|30
<3 20
10-
-0.1
0.0
0.1
0.2
0.3
0.4 0.5
Correlation
0.6
< CD O Q LU
O X _
5 z
1.0-
0.9
0.8-
0.7-
0.6-
I 0.5-
m
O fi A -
O 1
0.3-
02
01 -
0.0-
-0.1
93 090 091 091 091 092 OPS OS:
08
•1 080 091 095 087 090 OS6 089 06
001 0.84 0.82 089 087 087 0 SO 082 Ofl
0 83 0 85 0 65
A
B
C
D
E
F
G
H
-I
J
K
L
• M
N
50 100 150 200 250 300 350 400 450
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-130 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O$ in the Pittsburgh,
Pennsylvania, CSA.
3-194
-------
Salt Lake City CSA
25
€ 20-
8 1
o 10-
5-
-0.1 0.0 0.1 0.2
0.3
0.4 0.5
Correlation
00
O
1.0-
0.9-
0.8-
0.7-
0.6-
B 0.5-
1
O
0.4-
0.3-
0.2
0.1 -\
0.0
-0.1
•••
077 083 086 085 075 0.79 077 0.67
091 061 0.91 091 ''
0.78 0.86 0.85 0.84
83 081 071
0,79 077
7 0.64
°"
0.84 0.84 0.80
•
76 0.73 0.62
0.84 0.76 0.77 0.77 0.72
0.92 081 0.84 0.82 0.73
•A
B
C
D
-E
0.88 0.74 0.82 0.80 0.77
-H
50 100 150 200 250 300 350 400 450
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-131 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O$ in the Salt Lake City, Utah,
CSA.
3-195
-------
San Antonio CBSA
5-
c
i 3
O 2-
1
-0.1
0.0
0.1
0.2
0.3
0.4 0.5
Correlation
0.6
8
1.0
0.9-
0.8-
0.7
0.6 -\
I 0.5 -\
03
0.2-
0.1 -
0.0-
-0.1
-c
E
50 100 150 200 250 300 350 400
Distance (km)
450
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-132 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O$ in the San Antonio, Texas,
CBSA.
3-196
-------
San Francisco CSA
80-
40
20-
-c
.1 0.0 0.1 0.2 0
31
3 0
41
4 0
60
5 0
98
6 0
95
7 0
96
8 0
Correlation
•
< CO O Q
CO O O LLJ
50 100 150 200 250 300 350 400 450
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-133 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max Os in the San Francisco,
California, CSA.
3-197
-------
Seattle CSA
8-
s a-
g .
o 4
2-
1
| '
4
•5
Q
7
9
-0.1 0.0 0.1 0.2 0.3
0.4 0.5
Correlation
0.6 0.7
1.0-
09-
08-
0.7-
0.6
0.5-\
0.4
0.3-
0.2
0.1 -
0.0-
-0.1
0,59 064 056 0,57 042 0,10 062
079 080 074 049 032
076 081 080 066 025 071
084 061 026 0.65
080 059 038
0,58 025 0.62
0.18 059
0.24
50 100 150 200 250 300 350 400 450
Distance (km)
-J
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-134 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max O$ in the Seattle, Washington,
CSA.
3-198
-------
St. Louis CSA
50-
Ł 40-
3 -in.
0 20-
10-
29
60
30
-0.1
0.0
1.0-
09-
0.8-
0.7-
0.6-
04-
0.3-
0.2-
0.1 -
00-
-0.1
:.-i
01
0.2
0.3
O U Q 01 LL
0.4 0.5
Correlation
x
0.6
0.7
0.9
1.0
X -1 2 Z O Q.
^B 088 084
0.98 086 084 0.86 0.86 087
078
0.79 0.85 0.76
090 090 088
0 86 0 88 0 85
•A
B
C
•D
083 080 086 085 088 0.76 076 0,76 0.75
0.7B 079 0.77 0.77 0.77 0.87 Q.77 0.79 0.72
0.84 0.83 0.85 0.86 0.87 0.89 082 0.84 0.79
0.93 0.97 Oi)0 Oil
OSS ^H 085 076
0.76 0.86 0.63 0.63
0.76
0 78 0 82 0 BO
0 50 100 150 200 250 300 350 400 450
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of R.
Figure 3-135 Pair-wise monitor correlations expressed as a histogram (top),
contour matrix (middle) and scatter plot versus distance between
monitors (bottom) for 8-h daily max Os in the St. Louis, Missouri,
CSA.
3-199
-------
Atlanta CSA
30-
25-
c 20-
o 15-
0 10-
5-
22
33
0.00 0.05 0.10 0.15
0.20 025 030 0.35
Coefficient of Divergence
0.40 0.45 0.50 0.55
CO CJ
006 009 009 008 009 003 008 011 013 Oil
010 010 008 010 0.09 009 Oil 013 012
010 011 011 012 011 013 012 013
0.55-
0.50-
0.45-
0.40-
s
g 0.35-
Q 0.30-
"5
1 0.25-
0
8 0.20-
0.15-
0.10-
0.05-
n nn -
011 007 009 010 011 010 0 OB
011 008 005 010 013 012
OOB 010 010 011 007
006 0.07 012 010
010 013 012
013 Oil
Oil
. ..'V. '
• ***l " •
.(iCft
.«,*..
~A
-B
C
D
-E
-F
-G
-H
-1
•J
•K
* .*
•
50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-136 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max Q% in the
Atlanta, Georgia, CSA.
3-200
-------
Baltimore CSA
200-
•& 150-
3 100-
50-
7
231
138
2
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
Coefficient of Divergence
0.55
* IM • » •* •• •• i •:
i"» *•> •• *OT »• • » - " Ml «•*••*«• o
P» i« .'!•• iw >v *ar to» MI ** »«• *« *• •» * ••
0(* 0- JM »«• M* AW *W *« «» •" *M *"
.. ..^.. „„..„..„..„ .,.
1.. ^^f ,r .. ., .. .. ..
,. ,. .. „ .. .. .. «. „ .. ,.
tM ft «* «M «•» «>t
.. g ,. ,. . .
....... „ . .. .
H" ™ ** " *" "
.. .. .„
** IW OM
om
* «
* *fc *A * !• *
* * *jftKf*"*'» *"^ "
&$$$$ •' ' '
2r*^
•A
B
C
c
- b
„ c
p
G
- H
-1
•J
•K
-M
^N
-0
P
-s
•T
•U
V
w
X
•Y
-z
•AA
-AB
50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-137 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max O$ in the
Baltimore, Maryland, CSA.
3-201
-------
Birmingham CSA
30- 36
I 20"
0 10-
0.00 0.05 010 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55
Coefficient of Divergence
^ m ŁJ Q LU 1 1 f_5 T — j
i i i i i i « i i i
0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.11 0.08
009 0.09 0.09 0.10 0 Oe 0.09 0.11 0.03
009 009 009 009 008 Oil 009
0.55-
0.50-
0.45-
0.40-
s
g 0.35-
H>
i
5 0.30-
tt
1 025-
° 020-
o
0.15-
0.10-
0.05-
n nn
008 006 009 012 008
0.08 006 008 Oil 008
008 007 012 DOS
^B
0 08 012 0 07
010 0.08
010
*:*"*
. •*4*»t ** '
-A
-B
-C
-D
-E
- f-
-G
-H
-1
J
•
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-138 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max O$ in the
Birmingham, Alabama, CSA.
3-202
-------
Boston CSA
100-
- 80-
1 60-
0 40-
20-
2
75
114
18
1
000 005 0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40 0.45 050 055
O Q
u_ O I _
012 014 018 013 013 015 014 010 017 017 018 012 014 018 012 019 017 013 019 014
006 007 010 010 007 012 010 OOC 009 009 012 007 010 007 010 Oil 012 010 Oil
007 010 010 007 Oil 010 008 009 010 012 008 010 008 Oil Oil 012 0.11 012
010 011 007 012 008 008 007 009 013 Oil 010 Oil 010 009 013 010 014
0.07 0.08 0.08 0.05 012 0.11 0.10 0.09 012 011 0.11 0.15 0.12 0.10 0.14 0.11
008 008 007 012 012 011 008 012 012 0.11 015 013 009 015 010
0.55-
0.50-
0.45-
0,40-
|0.35-
1
Q 0.30 n
•s
I 0.25-
i
8 0.20-
0.15^
0.10-
0.05-
n nn -
009 008 007 010 009 010 009 009 008 012 011 010 012 010
009 012 013 012 008 012 012 Oil 016 013 007 016 009
008 0.09 0.10 010 0.09 0.11 0.12 0.11 0.10 0.12 0.11
Oil 010 013 006 P11 007 Oil Oil 013 012 012
0.09 0.15 012 0.11 0.13 0.09 0.08 0.15 0.09 0.15
014 012 008 012 012 010 013 012 013
0.12 0.12 0.10 0.17 014 0.05 017 0.07
01? 004 013 013 Oil 013 Oil
Oil 012 012 012 013 Oil
013 013 010 014 010
0.11 0.16 0.06 0.16
014 0.09 015
. " • • . • 0 16 0.06
. f»t . •. ;* . » . • •
\-fff'^-
-A
-B
-C
-D
-E
-F
-G
-H
•1
-J
-K
-L
-M
-N
-O
-P
-Q
-R
-s
-T
-U
50 100 150 200 250 300
Distance (km)
350 400 450
500
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-139 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max Os in the
Boston, Massachusetts, CSA.
3-203
-------
Chicago CSA
150-
= 100-
g
0 50-
2
159
163
1
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45
0.50
0.55
0.090.090.140.11 0030060 100100130130 120090140 10011 0.11 011 0.120.11 0.13 0.150 12 0.12 0.11 0.12
0.08 0 10 0.07 0.07 0.09 0 07 0.10 009 008 009 009 0 10 0.07 0.07 009 0.07 0.08 009 008 009 0 10 010 009 0.09
0.11 0090080090050 11 008009010011 010007008010009008010010011 0.090 100100.09
010011 014012013012009010013013011 010012010011 013010011 012012011 012
0.08011 0100 10011 0.070.070.09011 0.080.090.08 0.080.08009 009 0.090 10 0 10 0.08009
009009008011 008009010010008009008009009008010011 010011 010009
0.11 0.08011 013012011 013011 008011 0100120.10013014011 009012011
0.55-
0.50-
0.45-
0.40-
0)
I 0.35-
0)
fc
5 0.30-
•5
I 0.25-
Ł
3 0.20-
0.15-
0.10-
0.05-
n AD
0120 07 009 011 Oil 011 007008010009008012010011 010010010009
0.14013011 012012012011 011 011 013008013013012012013012
010011 013012008008011 009009013009010011 010012010
009011 Oil 0070.09009009007012008009011 0 11 009009
0.11 Oil 0090100100100100100090,10011 0 120,10010
0150.10010011 0.09012012010012013012009012
0.100.12011 0 120100130 13013011 0120.12011
0.08008008008011 008010009010008008
0 10 005 009 01 1 0.09 010010 003 010 0 10
010008011 011 012 009011 009009
0.09010008009011 009009010
011 009010008010009006
010010011 012012010
0.04011 012011 0.10
012013012010
0100.100.07
. . * % 011010
• '*-V?K."i'^**|v* •* OM
:**&$ftii&?* * *':'*'
-A
-B
-C
- D
-E
-F
-G
H
-1
J
-K
-1
-M
-N
-o
-P
-Q
-R
-S
-T
-u
-V
-w
-X
-Y
-z
>*{*•.*
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-140 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max O$ in the
Chicago, Illinois, CSA.
3-204
-------
Dallas CSA
80-
§60-
0 40-
20-
0.
99
64
1
7
1
XI 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
Coefficient of Divergence
Oil 010 010 009 0.14 016 0.08 009 010 016 016 012 015 012 012 0.10 013 Oil
0.07 0.07 0.09 0.08 0.08 O.It 0.08 0.06 009 008 0.09 0.09 010 0.09 0.09 010 013
009 007 009 Oil 00* 008 007 Oil 010 008 012 008 007 010 009 011
0.55-
0.50-
0.45-
0.40-
§ 0.35-
m
5 0.30-
•s
1 0.25-
8 0.20-
0.15-
0.10-
0.05-
n on -
'•"-" \ V
rtfwp^V-* '
• "
0.07 0.07 0.09 O.It 007 007 0.03 0.09 0.10 0.09 010 0.10 0.10 0.09 0.12
010 012 006 006 008 012 Oil 0.09 012 007 0.07 010 009 006
007 013 008 009 007 DM 010 0.09 010 0.11 013 O.QB 014
01S 012 006 006 008 010 006 013 011 012 011 016
0.08 0.10 O.fS 014 0.10 015 0.08 0.07 0.10 Oil 0.08
009 ON 010 0.10 0.12 0.08 0.09 0.10 008 009
010 0.09 006 009 010 008 007 0.09 0.12
007 011 008 013 012 013 010 015
010 009 011 Oil 013 007 Old
Oil 0.10 006 009 009 Oil
013 012 Oil 012 016
0.06 0.12 0.08 0.08
010 008 010
Oil 012
..
0.55
•A
-B
C
-D
E
-F
•G
-H
-1
-J
•K
•L
M
• N
-0
•P
•Q
•R
-s
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-141 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max Q% in the
Dallas, Texas, CSA.
3-205
-------
Denver CSA
60-
50-
| 40-
o 30-
0 20-
10-
1
66
10 14 7
j \ 1 , 1
0.00 0.05 010 015
0.20 0.25 030 035
Coefficient of Divergence
0.40 045 0.50 055
CD o
CD
008 013 008 008 Oil 009 010 012 010 010 010
0 16 019 019 021 020 022 022 0.25 022 0.10 019 021
Oil 008 009 009 009 010 012 008 010 017 010 010
008 009 008 009 009 010 012 009 017 0.05 010
0.55-
0.50-
0.45-
0,40-
g
|0.35-
1
Q 0.30 n
•s
I 0.25-
^S
0 02°"
0.15-
0.10-
0.05-
n nn -
0.05 007 005 007 007 008 008 017 005 007
006 005 005 007 007 006 019 0.06 007
007 005 008 008 007 018 007 006
0.06 0.06 0.07 0.06 0.20 0.06 0.07
^^1
O.OB 0.07 0.07 020 0.07 0.08
^W
0.09 0.05 0.22 0.08 008
008 008
019 006 007
"•'• . '
_•* » 017 019
* ... '
006
•• •
**f+ »*
-A
-B
•C
-D
-E
-F
-G
-H
•1
-J
-K
•L
-M
-N
-0
"fc^* i
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-142 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max Os in the
Denver, Colorado, CSA.
3-206
-------
Detroit CSA
20
I 15
5 10
5
25
000 005 0.10 0.15
020 025 0.30 0.35
Coefficient of Divergence
040 045 050 055
o
005 006 009 006 007 009 010 012
003 009 007 0.08 009 010 Oil
0.55-
0.50-
0.45-
0.40-
|0.35-
(| 0.30-
•5
I 025-
S 0.20-
015-
010-
0.05-
nnn
0.09 0.07 O.OS 008 009 Oil
0.08 0.10 0.10 Oil 0.13
0.08 0.08 011 0.12
0.08 0.07 0.11
^B
009 009
•
011
**•*•* •
• * * * * '
-A
-B
-C
-D
-E
-F
-G
-H
•I
50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-143 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max O$ in the
Detroit, Michigan, CSA.
3-207
-------
Houston CSA
0
o
100-
80-
60-
40-
20-
oc
22
114
56
18
1
0 0.05 010 015 0.20 0.25 030 035 0.40 045 0.50 0.55
< CO O Q
0.09 011 009
1
I
0
c
-------
Los Angeles CSA
400
„ 300-
c
O
100-
3
155
417
257
181
108
43 16 6 12 16
0.00 0.05 0 10 0.15 0.20 0.25 0.30 035 040 0.45 0.50 055
Coefficient of Divergence
M«aiMi!»»«<
qlnnuai»ai
0.55-
0.50-
0.45-
0.40-
S 0-35 -j
O 0.30-
B
! 0.25-I
O
0.20-
0.15-
0.10-
0.05-
0.00
1
I' •OVKWU^I*
•»:';)11 .-.'!(> 1>0 u .1: (M
i •«
I
I:
fj - . . ,'a»ni'ono
i
. :•'':•''. - '•*'•. ;-. •
^risr^i/-4:*- •'
•:Łs&f?Ł/V.J •»• .". "•»:
-/aKwij^r--^' vo1 :
.^^^;^y>;::... ;•:- . ~~
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-145 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max O$ in the
Los Angeles, California, CSA.
3-209
-------
Minneapolis CSA
20-
c IS-
C' -in -
O nu
5-
1
24
3
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55
Coefficient of Divergence
-------
New York CSA
250 i
_ 200
§ 150-
o 100
50-
8?
276
74
3
0.00 0.05 0,10 0.15 0.20 0.25 0.30 0.35 0.40 0,45 0.50
Coefficient of Divergence
0.55
<
<
CQ O Q
0.55-
050-
0.45-
0.40-
I 0.35-
5
5 030-
•5
1 0.25-
o
8 0,20-
0.15-
0.10-
0.05-
nnn
'" *" ' " '" '" " "" '" "" '" '" °" '"
' '
on M'
.V
on tit MI t«? i-i an ICH
: N I..NI
c
.M <•-
' -•'..'..•'* " .•••
• *J***^"5Ą«^% • > *.'V: • *
. . .^ » ^ c2vCTtir?"*v'^ '•* *
* * *^^S.?^*ti ^* ***••• v*" *•
*;:.^i^.'t- ""' '* : '
J
K
-L
-M
- N
-0
-p
-Q
- R
-S
-T
-u
-V
-w
-X
-Y
-z
-AA
-AB
-AC
-AD
50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-147 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max O3 in the
New York City, New York, CSA.
3-211
-------
Philadelphia CSA
so-
^ 60-
c
o 40 -
O
20-
o.oc
0.55-
0.50-
0.45-
0.40-
8
g 035-
15 0.30-
5
| 0.25-
1
o 02°-
0.15-
0.10-
0.05-
nnn
83
38
2 10 ;
0.05 010 015 020 0.25 0.30 035 040 0.45 0.50
Coefficient of Divergence
•
0.55
-A
-B
-c
- D
-E
-F
-G
- H
-I
-J
-K
L
M
N
-O
-P
-Q
50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-148 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max Os in the
Philadelphia, Pennsylvania, CSA.
3-212
-------
300
250
Ł 200
8 150
o 10Q
50
1 310
• h
Phoenix CBSA
0.00 005 010 015 020 0.25 030 0.35 0.40 0.45 0.50 055
Coefficient of Divergence
0.55-
0.50-
0.45-
0.40-
0.35-
5 030-
§ 0.25-
I
0.20-
0.15-
0.10-
0.05-
0.00
o
••::,
v*
j, • -'. •.
& :•*•*• '
A
-B
-C
-D
-E
-F
-G
-H
- I
J
-K
-L
-M
-N
-O
-P
Q
R
S
T
U
M
W
X
Y
Z
AA
AB
-AC
-AD
AE
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-149 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max O$ in the
Phoenix. Arizona, CBSA.
3-213
-------
Pittsburgh CSA
84
_60-
g 40-
O
20 1 6
0.00 0.05 0,10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0,50 0.55
Coefficient of Divergence
004 006 007 008 007 0.07 007 006 007 008 0.07 007 009
0.06 0.07 0,08 0.06 0.07 006 006 007 0.07 007 0.07 0.09
009 009 006 007 006 007 007 009 006 008 011
010 007 008 007 006 009 008 009 009 009
0.55-
050-
0.45-
0.40-
§ 0.35 -
5 0.30-
•5
1 0.25-
g
° 0.20-
0.15-
0.10-
0.05-
nnn
010 010 010 009 009 010 010 007 012
008 005 006 008 009 0.08 009 0.11
007 008 007 008 008 006 010
006 008 008 0.07 0.09 0.10
008 008 0.08 0.08 0,10
005 007 008 008
0.08 008 007
0.08 0 1 0
010
>*&/•
-A
-B
-C
-D
-E
-F
-G
- H
-1
-J
-K
- L
-M
-N
• "V*" *
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-150 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max O3 in the
Pittsburgh, Pennsylvania, CSA.
3-214
-------
Salt Lake City CSA
40
c 30 1
O 20 -j 13
10-^^
0.00 0.05 010 0.15 020 025 030 035
Coefficient of Divergence
tf CQ CJ O 1 1 1 u_ C5 T
i i i i i i t i
004 006 0.05 006 008 006 006
DOS 005 OO6 004 004 005
Q07 DOB 006 006 DOB
0.55-
0.50-
0.45-
0.40-
8
S °35'
O)
5
i5 0.30-
•5
1 0.25-
!Ł
° 0.20-
0.15-
0.10-
0.05-
n nn
0.07 0.07 0.06 0.06
O.OB 0 06 0.06
0 07 0.05
006
. ..
Łf*f*h '
0.40 0.45 050 055
— -5 ^ — I
0.07 007 0,07 008
007 0.06 007
0 07 0 OB 0 07 0 09
0 07 0.07 0,08
0.07 007 007 007
0.06 0.07 0.07 0.08
005 006 00$ 006
004 005 005
0 OS 0 05 0 06
V ^^_
0,05 005
^B
005
•A
B
-c
-D
-E
-F
-G
H
-I
-J
K
- L
50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-151 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max O$ in the
Salt Lake City, Utah, CSA.
3-215
-------
San Antonio CBSA
5-
c 4"
3 3-
0 2-
1 -
6
4
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Coefficient of Divergence
0.40
0.45
0.50
0.55
0.55-
0.50-
0.45-
0.40-
035-
0.30-
0.25-
0.20-
0.15-
0.10-
0.05-
0.00
A
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-152 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max O$ in the
San Antonio, Texas, CBSA.
3-216
-------
San Francisco CSA
150-
100-
50-
98
170
95
61
000 005 010 015 020 025 030 035 040 045 0.50
Coefficient of Divergence
055
> Iff • (.« : I *• ','! • • ; •• - t« IK Ml II! II.
..I ... M .- ^^^E> •.»•*•.» .
..-.,. nj . .. . .. „
+ !. 0.; *M *« 1- •• 411 t">
,,,
-»---•;• ;-••*
-
".. 1 "
' *
• *. * *
• *•• i .*••*..
.'••*'" •" .:.f; '•
^j?' v" ** *f * " * * • • ;
•'•*.^'>- -*?.,: •• "•*. .
>XS&"*&v.f •*
3^l$fi '''''•
.Ł* :~ • •'
.
-B
C
- E
F
-H
•J
•K
L
•M
N
•0
D
•Q
S
U
V
- w
X
•Y
•z
•AA
-AB
AC
AD
AE
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-153 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max O3 in the
San Francisco, California, CSA.
3-217
-------
Seattle CSA
o
15-
10-
5-
3
I
16
1C
7
1
1 . . . . .
0.00 005 010 015 020 025 030 035 040 0.45
Coefficient of Divergence
050 0.55
CD O
016 0.19 0.22 0.19 020 0.19 0.23 0.17 018
013 015 015 014 015 021 020 017
0.09 0.12 0.10 0.13 0.13 022 014
0.55-
0.50-
0.45-
0.40-
g 0.35-
1
Q 0.30-
•6
1 0.25-
0
fe
o 0.20-
0.15-
010-
0.05-
rinn -
Oil 009 014 016 023 017
008 014 014 020 0 IB
012 014 021 016
019 019 014
0.26 021
• « • p .
• • *
** . • ote
. . * *
. . *
* . • • .
• . .
*
-A
-B
-c
-D
-E
-F
-G
-H
-I
-J
50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-154 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max O$ in the
Seattle, Washington, CSA.
3-218
-------
St. Louis CSA
80-
Ł 60-
340-
20-
2
90
27
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45
Coefficient of Divergence
0.50
0.55
oaoauju.Oz_->*_i2zOa.
006 006 006 009 008 006 007 008 007 006 00« 010 009 010 010
009 009 0.09 010 008 010 010 0.08 0.08 008 010 009 0.11 0.10
O.OS 011 004 004 008 008 009 008 007 008 010 009 011
012 0.07 006 007 006 006 0.07 007 0.08 OOB 0.07 010
0.13 0.11 0.13 0.13 0.11 0.11 0.11 0.12 0.11 0.14 0.12
0.55-
0.50-
0.45-
0.40-
i 0.35-
0)
a)
3 0.30-
•5
| 0.25-
i|
8 020-
0.15-
010-
0.05-
n nn
0.08 008 009 009 008 009 009 008 Old
008 009 009 007 007 006 Oil 009 010
0.05 007 0.05 006 0.11 O.OS 0.07 0.08
007 005 007 010 006 005 008
0.06 0.07 010 0.08 0.09 0.09
0.06 010 0.06 008 007
mil
009 OOS O.oe 009
^^B
0.10 0.09 010
0.07 0.06
*
•
* # *
*• -* **\ « I *
4^*V j; * *
A
B
C
-D
•E
•F
G
-H
I
J
-K
•L
•M
-N
-0
P
0 50 100 150 200 250 300 350 400 450 500
Distance (km)
Note: The colors in the histogram bins correspond to the levels of the contour matrix. The histogram includes the number of monitor
pairs per bin and the contour matrix includes the numeric values of COD.
Figure 3-155 Pair-wise monitor coefficient of divergence expressed as a
histogram (top), contour matrix (middle) and scatter plot versus
distance between monitors (bottom) for 8-h daily max O$ in the
St. Louis, Missouri, CSA.
3-219
-------
3.9.4 Hourly Variations in O3 for the Urban Focus Cities
This section contains diel plots of 1-h avg O3 data to supplement the discussion on
hourly variations in O3 concentrations from Section 3.6.3.2 using data from the 20
urban focus cities first introduced in Section 3.6.2.1. Comparisons are made between
cold months (October-April) and warm months (May-September), using the year-
round data set, and between weekdays (Mon-Fri) and weekends (Sat-Sun) using the
warm-season data set.
3-220
-------
a
SIH
^ r?
50-
0-
0 (toys. 0 year-round *rte»
mean
median
=> 1*-«P
no yeaf-faund data
0 ctoys. 0 yew-round site*
no year-found data
Weekdays
327 nays. M y/3 rrr - sra
132d.lys.l1 Wjrm-s(r3*.-n
00.00 06:00 12.00 18:00 00.0000.00 05:00 12.00 18:00 0000 OO.OO 0600 12:00 18.00 00000000 08:00 1200 18:00 00:OC
hour hour hour hoir
Cold Months Warm Months Weekdays Weekends
637 days. 9 yea'- round sites
mean
median
= f-SS"
= I'-Mf*
9 days 9 year-round sites
327 flays. 28 warm-season srres 132 days. 26 vyarm -season sties
0000 OGOO 1200 18:00 00000000 06:00 12:00 1*00 CO00 00:00 0600 12:00 16.00 00000000 06.00 1203 18:00 COOC
Mur hour hour Mir
Warm Months
Weekdays
637 day&, t year-round site
— mean
327 days. fQwnfMHORtfRM 132 days. 10 waim-Maaon s»*s
00,00 06:00 1200 18:00 0000 0000 O&OO 12:00 16:00 0000 0000 0600 12:00 16:00 00000000 06.00 1200 16:00 OOCC
hour hour hour hour
Cold Months
Weekdays
637 days, 3 yvw-round sites
— mean
medtan
a Sf-SS"
' - ' ' 3S1
459 days 3 year-round sites
327 days, 21 warm-Mason arto
132 days, 21 warm -season stes
00.00 06CO 1200 td:00 0000 0000 0600 1200 !500 00.00 00:00 OSOO 12:00 1800 00000000 06.00 12-00 14:00 QQQC
hour hour hour hour
Note: No year-round monitors were available for the cold month/warm month comparison in the Atlanta CSA.
Figure 3-156 Diel patterns in 1-h avg O3 for select CSAs between 2007 and 2009
using the year-round data set for the cold month/warm month
comparison (left half) and the warm-season data set for the
weekday/weekend comparison (right half).
3-221
-------
Cold Months
Warm Months
Weekdays
150 -
100 -
so -
637 days. 11 year-round sites
«•« mean
median
<=> 5"-95"
459 days 11 year-round sites
327 days. 26 warm-season sites
132 days, 26 warm-season srtes
0000 06:00 12:00 ISOO
hour
Cold Months
000000:00 06-00 1200 18:00 00:0000:00 0600 12:00 18:00 000000:00 0600 1200 18:00 OOOC
hour hour Hour
Warm Months
Weekdays
35
150 -
100 -
so -
0 -
637 days. 19 year-round sites
— mean
median
c^ 5"-95"
459 days 19 year-round sites
327 days, 19 warm-season sites
132 days. 19 warm-season sites
0000 0600 12:00 1600
hour
Cold Months
000000:00 0600 1200 1600 00000000 0600 1200 18:00 000000:00 06:00 1200 1800 OOOC
hour hour hour
Warm Months
Weekdays
Weekends
637 days. 12 year-round sites
— mean
median
459 days 12 year-round sites 327 days, 15 warm-season sites 132 days, 15 warm-season sites
000000:00 06:00 1200 18:00 00:0000:00 0600 12:00 18:00 000000:00 0600 1200 1800 OOOC
hour hour hour
Cold Months
Warm Months
Weekdays
si««
o ~
0 days. 0 year-round sites
— mean
— median
c=^ 5"-95"
^^ 1*-98*
no year-round data
0 days, Q year-round sites
no year-round data
327 days, 9 warm-season srtes 132 days, 9 warm-season sites
0000 0600 12:00 I8OO
Hour
0000 00:00 06-00 1200 1600 00:00 OO'OO 0600 1200 1fl:00 0000 0000 0600 1200 1800 OOOC
hour hour hour
Note: No year-round monitors were available for the cold month/warm month comparison in the Detroit CSA.
Figure 3-157 Diel patterns in 1-h avg O3 for select CSAs between 2007 and 2009
using the year-round data set for the cold month/warm month
comparison (left half) and the warm-season data set for the
weekday/weekend comparison (right half).
3-222
-------
Cold Months
Warm Months
Weekdays
Weekends
150 -
§ a 10M
o s
|
5
| *M
537 days. 21 year-round sites
— mean
median
<=> 5*-95*
= l"-991n
459 days. 21 year-round sues
327 days, 21 warm-season siles
132 days. 21 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 0000 00:00 06:00 12:00 1800 00:00 00:00 0600 12:00 18:00 OOOC
hour hour hour hour
Cold Months
Warm Months
Weekdays
Weekends
o
!8 s
f I 100 H
6'
637 days, 47 year-round sites
mean
median
t=^ 5"-95"
= i" S3"
459 days, 47 year-round sites
327 days. 50 warm-season sites
132 days. 50 warm-season sites
00:00 06:00 12:00 18:00 00:0000:00 06:00 12:00 18:00 00:0000.00 06:00 1200 18:00 00:0000:00 06:00 12:00 18:00 OO.OC
riour hour hour hour
Cold Months
Warm Months
Weekdays
Weekends
O
| | too-
c
425 days. 2 year-round sites
— mean
median
306 days, 2 year-round sites
327 days, 8 warm-season sites
132 days, 8 warm-season sites
00:00 06:00 12:00 18:00 00:0000:00 06:00 12:00 18-00 00-0000:00 06:00 12:00 18:00 00:0000:00 06:00 12-00 18:00 00:OC
hour hour hour hour
Cold Months
Warm Months
Weekdays
Weekends
O
t
z
637 days. 20 year-round sites
• ••• mean
median
=> S" 95"
459 days, 20 year-round sites
327 days, 30 warm-season siles
132 days, 30 warm-season sites
00:00 06'QO 12:00 18:00 00:00 00:00 06:00 12:00 18.00 00.00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 1200 18:00 00:OC
hour hour hour hour
Figure 3-158 Diel patterns in 1-h avg O3 for select CSAs between 2007 and 2009
using the year-round data set for the cold month/warm month
comparison (left half) and the warm-season data set for the
weekday/weekend comparison (right half).
3-223
-------
Cold Months
Warm Months
Weekdays
Weekends
o
5*-95*
= 1--99"
459 days. 9 year-round sites
327 days, 17 warm-season siles
132 days. 17 warm-season sites
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 0000 00:00 06:00 12:00 1800 00:00 00:00 0600 12:00 18:00 OOOC
hour hour hour hour
Cold Months
Warm Months
Weekdays
Weekends
3
m
ll
637 days. 14 year-round sites
mean
median
r= 5"-95"
= i" S3"
459 days, 14 year-round sites
327 days. 31 warm-season sites
132 days. 31 warm-season sites
00:00 06:00 12:00 18:00 00:0000:00 06:00 12:00 18:00 00:0000.00 06:00 12:00 18:00 00:0000:00 06:00 12:00 18:00 OO.OC
tour hour hour hour
Cold Months
Warm Months
Weekdays
Weekends
o
I
.Q
637 days. 2 year-round sites
— mean
median
^=3 5*-95*
459 days, 2 year-round sites
327 days, 14 warm-season sites
132 days, 14 warm-season sites
00:00 06:00 12:00 18:00 00:0000:00 06:00 12:00 18-00 00-0000:00 06:00 12:00 18:00 00:0000:00 06:00 12-00 18:00 00:OC
hour hour hour hour
Cold Months
Warm Months
Weekdays
Weekends
o
>N
o
o
424 days. 2 year-round sites
• ••• mean
median
=> 5 95"
306 days, 2 year-round siles
327 days, 12 warm-season siles
132 days, 12 warm-season sites
00:00 06'00 12:00 18:00 00:00 00:00 06:00 12:00 18.00 00.00 00:00 06:00 12:00 18:00 00:00 00:00 06:00 1200 18:00 00:OC
hour hour hour hour
Figure 3-159 Diel patterns in 1-h avg O3 for select CSAs/CBSAs between 2007
and 2009 using the year-round data set for the cold month/warm
month comparison (left half) and the warm-season data set for the
weekday/weekend comparison (right half).
3-224
-------
Cold Months
Warm Months
Weekdays
Weekends
S
o
o
o
I
150 -
537 days. 5 year-round sites
— mean
median
<=> 5*-95*
= 1--99"
459 days. 5 year-round sites
327 days, 5 warm-season srtes
132 days. 5 warm-season srtes
00:00 06:00 12:00 18:00 00:00 00:00 06:00 12:00 18:00 0000 00:00 06:00 12:00 1800 00:00 00:00 0600 12:00 18:00 OO.OC
hour hour hour hour
Cold Months
Warm Months
Weekdays
Weekends
o
o
|
u
5
re
V)
637 days, 25 year-round sites
mean
median
t= 5"-95"
= i" S3"
459 days, 25 year-round sites
327 days. 31 warm-season sites
132 days. 31 warm-season sites
00:00 06:00 12:00 18:00 00:0000:00 06:00 12:00 18:00 00:0000:00 06:00 1200 18:00 00:0000:00 06:00 12:00 18:00 00:OC
hour hour hour hour
Cold Months
Warm Months
Weekdays
Weekends
O 5
1
ru
637 days. 5 year-round sites
— mean
median
^=3 5*-95*
459 days, 5 year-round srtes
327 days, 10 warm-season sites
132 days, 10 warm-season sites
00:00 06:00 12:00 18:00 00:0000:00 06:00 12:00 18-00 00-0000:00 06:00 12:00 18:00 00:0000:00 06:00 12-00 18:00 00:OC
hour hour hour hour
Cold Months
Warm Months
Weekdays
Weekends
O
in
O
635 days. 3 year-round sites
• -•- mean
median
= S" 95"
<= 1*-99™
i
459 days, 3 year-round sites
327 days, 16 warm-season sites
132 days, 16 warm-season sites
00:00 06'00 12:00 18:00 00:0000:00 06:00 12:00 1800 000000:00 06:00 12:00 18:00 00:0000:00 06:00 1200 18:00 00:OC
hour hour hour hour
Figure 3-160 Diel patterns in 1-h avg O3 for select CSAs/CBSAs between 2007
and 2009 using the year-round data set for the cold month/warm
month comparison (left half) and the warm-season data set for the
weekday/weekend comparison (right half).
3-225
-------
References
Acker. K; Febo. A; Trick. S: Perrino. C: Bruno. P; Wiesen. P; Moller; Wieprecht. W: Auel. R; Giusto. M;
Gever. A; Platt. U: Allegrini. I. (2006). Nitrous acid in the urban area of Rome. Atmos Environ 40: 3123-
3133. http://dx.doi.0rg/10.1016/i.atmosenv.2006.01.028
Andreae. MO. (1991). Biomass burning: its history, use, and distribution and its impact on environmental
quality and global climate. In JS Levine (Ed.), Global Biomass Burning: Atmospheric, Climatic, and
Biospheric Implications (pp. 1-21). Cambridge, MA: MIT Press.
Andreae. MO; Merlet. P. (2001). Emission of trace gases and aerosols from biomass burning. Global
Biogeochem Cycles 15: 955. http://dx.doi.org/10.1029/2000GB001382
Anton. M; Lopez. M; Vilaplana. JM; Kroon. M; McPeters. R; Bafion. M; Serrano. A. (2009). Validation of
OMI-TOMS and OMI-DOAS total ozone column using five Brewer spectroradiometers at the Iberian
peninsula. J Geophys Res 114: D14307. http://dx.doi.org/10.1029/2009JD012003
Appel. KW: Gilliland. A; Eder. B. (2005). An operational evaluation of the 2005 release of Models-3 CMAQ
version 45. Presentation presented at 4th Annual CMAS Models-3 Users' Conference, September 26-28,
2005, Chapel Hill, NC.
Archibald. AT; Cooke. MC: Utembe. SR; Shallcross. DE; Dement. RG: Jenkin. ME. (2010). Impacts of
mechanistic changes on HOx formation and recycling in the oxidation of isoprene. Atmos Chem Phys 10:
8097-8118. http://dx.doi.org/10.5194/acp-10-8097-2010
Archibald. AT; Levine. JG: Abraham. NL; Cooke. MC: Edwards. PM; Heard. DE; Jenkin. ME; Karunaharan.
A; Pike. RC: Monks. PS: Shallcross. DE; Telford. PJ: Whallev. LK; Pyle. JA. (2011). Impacts of HOx
regeneration and recycling in the oxidation of isoprene: Consequences for the composition of past, present
and future atmospheres. Geophys Res Lett 38: L05804. http://dx.doi.org/10.1029/2010GL046520
Arnold. JR; Dennis. RL; Tonnesen. GS. (2003). Diagnostic evaluation of numerical air quality models with
specialized ambient observations: testing the Community Multiscale Air Quality modeling system (CMAQ)
at selected SOS 95 ground sites. Atmos Environ 37: 1185-1198.
Arshinov. MY; Belan. BD; Krasnov. OA; Kovalevskii. VK; Pirogov. VA; Plotnikov. AP; Tolmachev. GN;
Fofonov. AV. (2002). Comparison of ultraviolet and chemiluminescent ozonometers. Atmos Ocean 15: 656-
658.
Arunachalam. S. (2009). Peer review of source apportionment tools in CAMx and CMAQ. (Contract No EP-
D-07-102). University of North Carolina Institute for the Environment, prepared for US Environmental
Protection Agency. http://www.epa.gov/scram001/reports/SourceApportionmentPeerReview.pdf
ATMET (Atmospheric, Meteorological, and Environmental Technologies). (2011). Atmospheric,
meteorological, and environmental technologies. Available online at http://atmet.com/ (accessed January
28,2011).
Barrie. LA; Bottenheim. JW: Schnell. RC: Crutzen. PJ: Rasmussen. RA. (1988). Ozone destruction and
photochemical reactions at polar sunrise in the lower Arctic atmosphere. Nature 334: 138-141.
Beckerman. B; Jerrett. M; Brook. JR; Verma. DK; Arain. MA; Finkelstein. MM. (2008). Correlation of
nitrogen dioxide with other traffic pollutants near a major expressway. Atmos Environ 42: 275-290.
http://dx.doi.0rg/10.1016/i.atmosenv.2007.09.042
Beer. R. (2006). TES on the aura mission: Scientific objectives, measurements, and analysis overview. IEEE
Trans Geosci Remote Sens 44: 1102-1105. http://dx.doi.org/10.1109/TGRS.2005.863716
Berkowitz. CM; Fast. JD; Sprinston. SR: Larsen. RJ; Spicer. CW: Doskev. PV: Hubbe. JM; Plastridge. R.
(1998). Formation mechanisms and chemical characteristics of elevated photochemical layers over the
northeast United States. J Geophys Res 103: 10,631-610,647.
3-226
-------
Berkowitz. CM; Shaw. WJ. (1997). Airborne measurements of boundary layer chemistry during the Southern
Oxidant Study: A case study. J Geophys Res 102: 12,795-712,804. http://dx.doi.org/10.1029/97JD00417
Binkowski. F; Roselle. S. (2003). Models-3 Community Multiscale Air Quality(CMAQ) model aerosol
component 1. Model description. J Geophys Res 108: 4183. http://dx.doi.org/10.1029/2001JD001409
Binkowski. FS: Arunachalam. S: Adelman. Z; Pinto. JR (2007). Examining photolysis rates with a prototype
online photolysis module in CMAQ. J Appl Meteor Climatol 46: 1252-1256.
Bishop. GA; Stedman. DH. (2008). A decade of on-road emissions measurements. Environ Sci Technol 42:
1651-1656. http://dx.doi.org/10.1021/es702413b
Bloomer. BJ; Stehr. JW: Piety. CA; Salawitch. RJ; Dickerson. RR. (2009). Observed relationships of ozone air
pollution with temperature and emissions. Geophys Res Lett 36: L09803.
http://dx.doi.org/10.1029/2009GL037308
Blumenthal. PL; Lurmann. FW: Kumar. N; Dye. TS: Ray. SE; Korc. ME; Londergan. R; Moore. G. (1997).
Transport and mixing phenomena related to ozone exceedances in the northeast US (analysis based on
NARSTO-northeast data). Santa Rosa, CA: Sonoma Technology.
http://capita.wustl.edu/otag/reports/otagrept/otagrept.html
Boersma. KF; Eskes. HJ; Brinksma. EJ. (2004). Error analysis for tropospheric NO2 retrieval from space. J
Geophys Res 109: D04311. http://dx.doi.org/10.1029/2003JD003962
Bonn. B; Von Kuhlmann. R; Lawrence. MG. (2004). High contribution of biogenic hydroperoxides to
secondary organic aerosol formation. Geophys Res Lett 31: L10108.
http://dx.doi.org/10.1029/2003GL019172
Brodin. M; Helmig. D; Oltmans. S. (2010). Seasonal ozone behavior along an elevation gradient in the
Colorado Front Range Mountains. Atmos Environ 44: 5305-5315.
http://dx.doi.0rg/10.1016/i.atmosenv.2010.06.033
Burgard. DA; Bishop. GA; Stedman. DH; Gessner. VH; Daeschlein. C. (2006). Remote sensing of in-use
heavy-duty diesel trucks. Environ Sci Technol 40: 6938-6942. http://dx.doi.org/10.1021/es060989a
Burlev. JD; Ray. JD. (2007). Surface ozone in Yosemite National Park. Atmos Environ 41: 6048-6062.
Buzica. D; Gerboles. M; Plaisance. H. (2008). The equivalence of diffusive samplers to reference methods for
monitoring O3, benzene and NO2 in ambient air. J Environ Monit 10: 1052-1059.
http://dx.doi.org/10.1039/b802260g
Byun. D; Schere. KL. (2006). Review of the governing equations, computational algorithms, and other
components of the models-3 community multiscale air quality (CMAQ) modeling system [Review]. Appl
Mech Rev 59: 51-77. http://dx.doi.Org/10.1115/l.2128636
Byun. DW: Ching. JKS. (1999). Science algorithms of the EPA models-3 community multiscale air quality
(CMAQ) modeling system. (EPA/600-R-99-030). Washington, DC: U.S. Environmental Protection Agency.
http://www.epa.gov/asmdnerl/CMAQ/CMAQscienceDoc.html
Carter. WPL. (1995). Computer modeling of environmental chamber studies of maximum incremental
reactivities of volatile organic compounds. Atmos Environ 29: 2513-2527.
Carter. WPL; Seinfeld. JH. (2012). Winter ozone formation and VOC incremental reactivities in the Upper
Green River Basin of Wyoming. Atmos Environ 50: 255-266.
http://dx.doi.0rg/10.1016/i.atmosenv.2011.12.025
Chan. E; Vet. RJ. (2010). Baseline levels and trends of ground level ozone in Canada and the United States.
Atmos Chem Phys 10: 8629-8647. http://dx.doi.org/10.5194/acp-10-8629-2010
Chen. X; Hopke. PK; Carter. WP. (2011). Secondary organic aerosol from ozonolysis of biogenic volatile
organic compounds: Chamber studies of particle and reactive oxygen species formation. Environ Sci
Technol 45: 276-282. http://dx.doi.org/10.1021/esl02166c
3-227
-------
Ching. J; Herwehe. J: Swall. J. (2006). On joint deterministic grid modeling and sub-grid variability
conceptual framework for model evaluation. Atmos Environ 40: 4935-4945.
Civerolo. KL; Mao. HT; Rao. ST. (2003). The airshed for ozone and fine paniculate pollution in the eastern
United States. Pure Appl Geophys 160: 81-105.
Conrad. R; Seiler. W. (1985). Influence of temperature, moisture, and organic carbon on the flux of H2 and CO
between soil and atmosphere: Field studies in subtropical regions. J Geophys Res 90: 5699-5709.
Cooper. OR; Oltmans. SJ: Johnson. BJ; Brioude. J: Angevine. W: Trainer. M; Parrish. DP; Ryerson. TR;
Pollack. I; Cullis. PD; Ives. MA; Tarasick. DW: Al-Saadi. J: Stainer. I. (2011). Measurement of western
U.S. baseline ozone from the surface to the tropopause and assessment of downwind impact regions. J
Geophys Res 116: 1-22.
Cooper. OR; Parrish. DP; Stohl. A; Trainer. M; Nedelec. P; Thouret. V: Gammas. JP; Oltmans. SJ: Johnson.
BJ; Tarasick. D; Leblanc. T; Mcdermid. IS: Jaffe. D; Gao. R; Stith. J: Ryerson. T; Aikin. K; Campos. T;
Weinheimer. A; Avery. MA. (2010). Increasing springtime ozone mixing ratios in the free troposphere over
western North America. Nature 463: 344-348. http://dx.doi.org/10.1038/nature08708
Corsmeier. U; Kalthhoff. N; Kolle. O: Motzian. M; Fiedler. F. (1997). Ozone concentration jump in the stable
nocturnal boundary layer during a LLJ-event. Atmos Environ 31: 1977-1989.
Crounse. JD; Paulot. F; Kjaergaard. HG: Wennberg. PO. (2011). Peroxy radical isomerization in the oxidation
of isoprene. Phys Chem Chem Phys 13: 13607-13613. http://dx.doi.org/10.1039/clcp21330i
D'Anna. B; Jammoul. A; George. C: Stemmler. K; Fahrni. S: Ammann. M; Wisthaler. A. (2009). Lightinduced
ozone depletion by humic acid films and submicron aerosol particles. J Geophys Res 114: D12301.
http://dx.doi.org/10.1029/2008JD011237
Dallmann. TR; Harlev. RA. (2010). Evaluation of mobile source emission trends in the United States. J
Geophys Res 115: D14305. http://dx.doi.org/10.1029/2010JD013862
de Gouw. JA: Brock. CA: Atlas. EL: Bates. TS: Fehsenfeld. FC: Goldan. PD: JS. H: Kuster. WC: Lerner. BM:
Matthew. BM; Middlebrook. AM; Onasch. TB; Peltier. RE; Quinn. PK; Senff. CJ: Stohl. A; Sullivan. AP;
Trainer. M; Warneke. C: Weber. RJ; Williams. EJ. (2008). Sources of particulate matter in the northeastern
United States in summer: 1. Direct emissions and secondary formation of organic matter in urban plumes. J
Geophys Res 113: D08301.
Dickerson. RR; Rhoads. KP; Carsev. TP; Oltmans. SJ: Burrows. JP: Crutzen. PJ. (1999). Ozone in the remote
marine boundary layer: A possible role for halogens. J Geophys Res 104: 21,385-321,395.
Dochertv. KS: Wu. W: Lim. YB; Ziemann. PJ. (2005). Contributions of organic peroxides to secondary aerosol
formed from reactions of monoterpenes with O3. Environ Sci Technol 39: 4049-4059.
http://dx.doi.org/10.1021/es050228s
Doyle. M; Sexton. KG: Jeffries. H; Bridge. K; Jaspers. I. (2004). Effects of 1,3-butadiene, isoprene, and their
photochemical degradation products on human lung cells. Environ Health Perspect 112: 1488-1495.
http://dx.doi.org/10.1289/ehp.7022
Doyle. M; Sexton. KG: Jeffries. H; Jaspers. I. (2007). Atmospheric photochemical transformations enhance
1,3-butadiene-induced inflammatory responses in human epithelial cells: The role of ozone and other
photochemical degradation products. Chem Biol Interact 166: 163-169.
http://dx.doi.0rg/10.1016/i.cbi.2006.05.016
Duncan. BN; Yoshida. Y; Olson. JR; Sillman. S: Martin. RV; Lamsal. L; Hu. Y; Pickering. KE; Retscher. C:
Allen. DJ. (2010). Application of OMI observations to a space-based indicator of NOx and VOC controls
on surface ozone formation. Atmos Environ 44: 2213-2223.
http://dx.doi.0rg/10.1016/i.atmosenv.2010.03.010
Dunker. AM; Yarwood. G: Ortmann. JP; Wilson. GM. (2002). Comparison of Source Apportionment and
Source Sensitivity of Ozone in a Three-Dimensional Air Quality Model. Environ Sci Technol 36: 2953-
2964. http://dx.doi.org/10.1021/es011418f
3-228
-------
Dunlea. EJ; Herndon. SC: Nelson. DP; Volkamer. RM; Lamb. BK; Allwine. EJ; Grutter. M; Ramos Villegas.
CR: Marquez. C: Blanco. S: Cardenas. B; Kolb. CE; Molina. LT; Molina. MJ. (2006). Technical note:
Evaluation of standard ultraviolet absorption ozone monitors in a polluted urban environment. Atmos Chem
Phys Discuss 6: 2241-2279.
Ebeling. D; Patel. V: Findlav. M; Stetter. J. (2009). Electrochemical ozone sensor and instrument with
characterization of the electrode and gas flow effects. Sens Actuators B 137: 129-133.
http://dx.doi.0rg/10.1016/i.snb.2008.10.038
Eder. B; Yu. S. (2005). A performance evaluation of the 2004 release of Models-3 CMAQ. Atmos Environ 40:
4811-4824. http://dx.doi.0rg/10.1016/j.atmosenv.2005.08.045
Eisele. FL; Mount. GH; Tanner. D; Jefferson. A; Shetter. R; Harder. JW: Williams. EJ. (1997). Understanding
the production and interconversion of the hydroxyl radical during the tropospheric OH photochemistry
experiment. J Geophys Res 102: 6457-6465. http://dx.doi.org/10.1029/96JD02207
Emery. C: Jung. J; Downey. N; Johnson. J; Jimenez. M; Yarvvood. G: Morris. R. (2012). Regional and global
modeling estimates of policy relevant background ozone over the United States. Atmos Environ 47: 206-
217. http://dx.doi.0rg/10.1016/i.atmosenv.2011.ll.012
Emmerson. KM; Evans. MJ. (2009). Comparison of tropospheric gas-phase chemistry schemes for use within
global models. Atmos Chem Phys 9: 1831-1845. http://dx.doi.org/10.5194/acpd-8-19957-2008
ENVIRON (ENVIRON Holdings Inc.). (2005). CAMx. Available online at http://www.camx.com/over/
(accessed January 28, 2011).
Fang. Y; Fiore. AM; Horowitz. LW: Lew II. H; Hu. Y; Russell. AG. (2010). Sensitivity of the NOy budget
over the United States to anthropogenic and lightning NOx in summer. J Geophys Res 115: D18312.
http://dx.doi.org/10.1029/2010JD014079
Fehsenfeld. FC: Ancellet. G: Bates. TS: Goldstein. AH; Hardestv. RM; Honrath. R; Law. KS: Lewis. AC:
Leaitch. R; McKeen. S: Meagher. J: Parrish. DP; Pszenny. AAP; Russell. PB; Schlager. H; Seinfeld. J:
Talbot R: Zbinden. R. (2006). International consortium for atmospheric research on transport and
transformation (ICARTT): North America to Europe: Overview of the 2004 summer field study. J Geophys
Res 111: D23S01.21-D23S01.36. http://dx.doi.org/10.1029/2006JD007829
Fehsenfeld. FC: Trainer. M; Parrish. DP; Volz-Thomas. A; Penkett. S. (1996). North Atlantic Regional
Experiment (NARE) 1993 summer intensive: Foreword. J Geophys Res 101: 28869-28875.
Finlavson-Pitts. BJ; Pitts. JN. Jr. (1986). Atmospheric chemistry: Fundamentals and experimental techniques.
New York, NY: John Wiley & Sons.
Fiore. A; Dentener. F; Wild. O: Cuvelier. C: Schultz. M; Hess. P; Textor. C: Schulz. M; Dohertv. R: Horowitz.
L; MacKenzie. I; Sanderson. M; Shindell. D: Stevenson. D; Szopa. S: Van Dingenen. R; Zeng. G: Atherton.
C: Bergmann. D; Bey. I; Carmichael. G: Collins. W: Duncan. B; Faluvegi. G: Folberth. G: Gauss. M;
Gong. S: Hauglustaine. D: Hollowav. T; Isaksen. I; Jacob. D: Jonson. J: Kaminski. J: Keating. T; Lupu. A;
Marnier. E; Montanaro. V; Park. R; Pitari. G: Pringle. K; Pyle. J: Schroeder. S: Vivanco. M; Wind. P;
Wojcik. G: Wu. S: Zuber. A. (2009). Multimodel estimates of intercontinental source-receptor relationships
for ozone pollution. J Geophys Res 114: D04301. http://dx.doi.org/10.1029/2008JD010816
Fiore. A; Jacob. DJ: Liu. H; Yantosca. RM: Fairlie. TD; Li. Q. (2003). Variability in surface ozone background
over the United States: Implications for air quality policy. J Geophys Res 108: 4787.
http://dx.doi.org/10.1029/2003JD003855
Fiore. AM; Horowitz. LW: Purves. DW: Lew. H. II; Evans. MJ: Wang. Y; Li. Q; Yantosca. RM. (2005).
Evaluating the contribution of changes in isoprene emissions to surface ozone trends over the eastern
United States. J Geophys Res 110: D12303. http://dx.doi.org/10.1029/2004JD005485
Fischer. EV (2004). Summertime ozone at Mount Washington: Meteorological controls at the highest peak in
the northeast. J Geophys Res 109: D24303. http://dx.doi.org/10.1029/2004JD004841
3-229
-------
Fuentes. JD; Wang. D; Bowling. PR; Potosnak. M; Monson. RK; Goliff. WS: Stockwell. WR. (2007).
Biogenic hydrocarbon chemistry within and above a mixed deciduous forest. J Atmos Chem 56: 165-185.
http://dx.doi.org/10.1007/sl0874-006-9048-4
Fuentes. M; Raftery. AE. (2005). Model evaluation and spatial interpolation by Bayesian combination of
observations with outputs from numerical models. Biometrics 61: 36-45.
Fusco. AC: Logan. JA. (2003). Analysis of 1970-1995 trends in tropospheric ozone at Northern Hemisphere
midlatitudes with the GEOS-CHEM model. J Geophys Res 108: 4449.
http://dx.doi.org/10.1029/2002JD002742
Gavdos. TM; Finder. R; Koo. B; Fahev. KM; Yarwood. G: Pandis. SN. (2007). Development and application of
a three-dimensional aerosol chemical transport model, PMCAMx. Atmos Environ 41: 2594-2611.
http://dx.doi.0rg/10.1016/i.atmosenv.2006.ll.034
Generoso. S: Bey. I; Attie. JL; Breon. FM. (2007). A satellite- and model-based assessment of the 2003
Russian fires: Impact on the arctic region. J Geophys Res 112: 5302.
Gilliland. AB; Hogrefe. C: Finder. RW: Godowitch. JM; Folev. KL; Rao. ST. (2008). Dynamic evaluation of
regional air quality models: Assessing changes in O3 stemming from changes in emissions and
meteorology. Atmos Environ 42: 5110-5123. http://dx.doi.Org/10.1016/i.atmosenv.2008.02.018
Godowitch. JM; Gilliland. AB; Draxler. RR; Rao. ST. (2008). Modeling assessment of point source NOx
emission reductions on ozone air quality in the eastern United States. Atmos Environ 42: 87-100.
http://dx.doi.0rg/10.1016/i.atmosenv.2007.09.032
Goldstein. A; Galbally. I. (2007). Known and unexplored organic constituents in the earths atmosphere.
Environ Sci Technol 41:15141521. http://dx.doi.org/10.1021/es072476p
Goldstein. AH; Millet. DB; McKay. M; Jaegle. L; Horowitz. L; Cooper. O; Hudman. R; Jacob. DJ: Oltmans. S:
Clarke. A. (2004). Impact of Asian emissions on observations at Trinidad Head, California, during ITCT
2K2. J Geophys Res 109: D23S17. http://dx.doi.org/10.1029/2003JD004406
Gottardini. E; Cristofori. A; Cristofolini. F; Ferretti. M. (2010). Variability of ozone concentration in a montane
environment, northern Italy. Atmos Environ 44: 147-152. http://dx.doi.Org/10.1016/j.atmosenv.2009.10.017
Greenberg. JP; Guenther. AB: Turnipseed. A. (2009). Tethered balloon-based soundings of ozone, aerosols, and
solar radiation near Mexico City during MIRAGE-MEX. Atmos Environ 43: 2672-2677.
http://dx.doi.0rg/10.1016/i.atmosenv.2009.02.019
Grell. GA; Emeis. S: Stockwell. WR: Schoenemever. T; Forkel. R: Michalakes. J: Knoche. R: Seidl. W.
(2000). Application of a multiscale, coupled MM5/chemistry model to the complex terrain of the VOTALP
valley campaign. Atmos Environ 34: 1435-1453.
Guenther. A; Geron. C: Pierce. T; Lamb. B; Harlev. P; Fall. R. (2000). Natural emissions of non-methane
volatile organic compounds, carbon monoxide, and oxides of nitrogen from North America. Atmos Environ
34: 2205-2230. http://dx.doi.org/10.1016/S1352-2310(99)00465-3
Guenther. A; Karl. T; Harlev. P; Wiedinmyer. C: Palmer. PI; Geron. C. (2006). Estimates of global terrestrial
isoprene emissions using MEGAN (Model of Emissions of Gases and Aerosols from Nature). Atmos Chem
Phys 6: 3181-3210. http://dx.doi.org/10.5194/acp-6-3181-2006
Hains. JC: Taubman. BF; Thompson. AM; Stehr. JW: Marufu. IT; Doddridge. BG: Dickerson. RR. (2008).
Origins of chemical pollution derived from Mid-Atlantic aircraft profiles using a clustering technique.
Atmos Environ 42: 1727-1741. http://dx.doi.Org/10.1016/i.atmosenv.2007.ll.052
Hameed. S: Pinto. JP: Stewart. RW. (1979). Sensitivity of the predicted CO-OH-CH4 perturbation to
tropospheric NOx concentrations. J Geophys Res 84: 763-768.
Harlev. RA; Marr. LC: Lehner. JK; Giddings. SN. (2005). Changes in motor vehicle emissions on diurnal to
decadal time scales and effects on atmospheric composition. Environ Sci Technol 39: 5356-5362.
Harvard University. (2010a). GEOSChem Model [Computer Program]. Cambridge, MA. Retrieved from
http://acmg.seas.harvard.edu/geos/
3-230
-------
Harvard University. (201 Ob). GEOSChem Overview. Available online at
http://acmg.seas.harvard.edu/geos/geos overview.html (accessed January 28, 2011).
Henderson. BH: Finder. RW: Crooks. J: Cohen. RC: Hutzell. WT: Sarwar. G: Goliff. WS: Stockwell. WR:
Fahr. A; Mathur. R; Carlton. AG: Vizuete. W. (2011). Evaluation of simulated photochemical partitioning of
oxidized nitrogen in the upper troposphere. Atmos Chem Phys 11: 275-291. http://dx.doi.Org/10.5194/acp-
11-275-2011
Hess. GD; Carnovale. F; Cope. ME; Johnson. GM. (1992). The evaluation of some photochemical smog
reaction mechanisms!. Temperature and initial composition effects. Atmos Environ 26: 625-641.
http://dx.doi.org/10.1016/0960-1686(92)90174-J
Hocking. WK; Carey-Smith. T; Tarasick. DW: Argall. PS: Strong. K; Rochon. Y; Zawadzki. I; Taylor. PA.
(2007). Detection of stratospheric ozone intrusions by windprofiler radars. Nature 450: 281-284.
http://dx.doi.org/10.1038/nature06312
Hofzumahaus. A; Rohrer. F; Keding. L; Bohn. B; Brauers. T; Chih-Chung. C: Fuchs. H; Holland. F; Kita. K;
Kondo. Y; Xin. L; Shengrong. L; Min. S: Limin. Z; Wahner. A; Yuanhang. Z. (2009). Amplified trace gas
removal in the troposphere. Science 324: 1702-1704. http://dx.doi.org/10.1126/science.1164566
Horowitz. LW: Fiore. AM; Milly. GP; Cohen. RC: Perring. A; Wooldridge. PJ: Hess. PG: Emmons. LK;
Lamarque. JF. (2007). Observational constraints on the chemistry of isoprene nitrates over the eastern
United States. J Geophys Res 112: D12S08. http://dx.doi.org/10.1029/2006JD007747
Hoskins. BJ. (1972). Non-boussinesq effects and further development in a model of upper tropospheric
frontogenesis. Q J Roy Meteorol Soc 98: 532-541. http://dx.doi.org/10.1002/qi.49709841705
Hsu. J: Prather. MJ. (2009). Stratospheric variability and tropospheric ozone. J Geophys Res 114: D06102.
http://dx.doi.org/10.1029/2008JD010942
Hudman. RC: Jacob. DJ: Turquetv. S: Leibensperger. EM; Murray. IT; Wu. S: Gilliland. AB; Avery. M;
Bertram. TH; Brune. W: Cohen. RC: Dibb. JE; Flocke. FM; Fried. A; Hollowav. J: Neuman. JA; Orville. R;
Perring. A; Ren. X; Sachse. GW: Singh. HB; Swanson. A; Wooldridge. PJ. (2007). Surface and lightning
sources of nitrogen oxides over the United States: Magnitudes, chemical evolution, and outflow. J Geophys
Res 112: D12S05. http://dx.doi.org/10.1029/2006JD007912
Hudman. RC: Murray. IT; Jacob. DJ: Millet. DB; Turquetv. S: Wu. S: Blake. PR; Goldstein. AH; Hollowav. J:
Sachse. GW. (2008). Biogenic versus anthropogenic sources of CO in the United States. Geophys Res Lett
35: L04801. http://dx.doi.org/10.1029/2007gl032393
Husar. RB; Renard. WP. (1998). Ozone as a function of local wind speed and direction: Evidence of local and
regional transport. 91st annual meeting and exhibition of the Air & Waste Management Association, San
Diego, CA.
Inman. RE; Ingersoll. RB; Lew. EA. (1971). Soil: A natural sink for carbon monoxide. Science 172: 1229-
1231. http://dx.doi.org/10.1126/science.172.3989.1229
Ito. A; Sillman. S: Penner. JE. (2009). Global chemical transport model study of ozone response to changes in
chemical kinetics and biogenic volatile organic compounds emissions due to increasing temperatures:
sensitivities to isoprene nitrate chemistry and grid resolution. J Geophys Res 114: D09301.
http://dx.doi.org/10.1029/2008id011254
Jacob. DJ. (1999). Introduction to atmospheric chemistry. New Jersey: Princeton University Press.
Jacob. DJ: Horowitz. LW: Munger. JW: Heikes. BG: Dickerson. RR; Artz. RS: Keene. WC. (1995). Seasonal
transition from NOx- to hydrocarbon-limited conditions for ozone production over the eastern United
States in September. J Geophys Res 100: 9315-9324.
Jacobson. MZ. (2002). Atmospheric pollution: History, science, and regulation. New York: Cambridge
University Press.
Jacobson. MZ. (2005). Fundamentals of atmospheric modeling (2 ed.). New York: Cambridge University
Press.
3-231
-------
Jaegle. L; Jacob. DJ: Brune. WH; Wennberg. PO. (2001). Chemistry of HOx radicals in the upper troposphere.
Atmos Environ 35: 469-489. http://dx.doi.org/10.1016/81352-2310(00)00376-9
Jaffe. D; Chand. D; Hafner. W: Westerling. A; Spracklen. D. (2008). Influence of fires on O-3 concentrations
in the western US. Environ Sci Technol 42: 5885-5891. http://dx.doi.org/10.1021/es800084k
Jaffe. D: Price. H; Parrish. D; Goldstein. A; Harris. J. (2003). Increasing background ozone during spring on
the west coast of North America. Geophys Res Lett 30: 1613. http://dx.doi.org/10.1029/2003GL017024
Jaffe. D: Ray. J. (2007). Increase in surface ozone at rural sites in the western US. Atmos Environ 41: 5452-
5463.
Jaffe. DA; Wigder. NL. (2012). Ozone production from wildfires: A critical review. Atmos Environ 51: 1-10.
http://dx.doi.0rg/10.1016/i.atmosenv.2011.ll.063
James. P; Stohl. A; Forster. C: Eckhardt S: Seibert. P; Frank. A. (2003). A 15-year climatology of stratosphere-
troposphere exchange with a Lagrangian particle dispersion model: 2. Mean climate and seasonal
variability. J Geophys Res 108: D12. http://dx.doi.org/10.1029/2002JD002639
Jimenez. JL; Jayne. JT; Shi. Q; Kolb. CE; Worsnop. PR; Yourshaw. I; Seinfeld. JH; Flagan. RC: Zhang. X;
Smith. KA. (2003). Ambient aerosol sampling using the Aerodyne Aerosol Mass Spectrometer. J Geophys
Res 108:8425.
Jo. WK; Park. JH. (2005). Characteristics of roadside air pollution in Korean metropolitan city (Daegu) over
last 5 to 6 years: Temporal variations, standard exceedances, and dependence on meteorological conditions.
Chemosphere 59: 1557-1573. http://dx.doi.Org/10.1016/i.chemosphere.2004.12.021
Johnson. D; Jenkin. ME; Wirtz. K; Martin-Riviejo. M. (2004). Simulating the formation of secondary organic
aerosol from the photooxidation of toluene. Environ Chem 1: 150-165.
Johnson. TR. (1995). Recent advances in the estimation of population exposure to mobile source pollutants. J
Expo Sci Environ Epidemiol 5: 551-571.
Kang. D; Aneja. VP; Matfaur. R; Ray. JD. (2003). Nonmethane hydrocarbons and ozone in three rural southeast
United States national parks: A model sensitivity analysis and comparison to measurements. J Geophys Res
108: 4604. http://dx.doi.org/10.1029/2002JD003054
Kasibhatla. P: Chameides. WL. (2000). Seasonal modeling of regional ozone pollution in the eastern United
States. Geophys Res Lett 27: 1415-1418. http://dx.doi.org/10.1029/1999GL011147
Kavnak. B; Hu. Y; Martin. RV; Russell. AG: Choi. Y; Wang. Y. (2008). The effect of lightning NOx
production on surface ozone in the continental United States. Atmos Chem Phys 8: 5151-5159.
Kim. SW: Heckel. A; Mckeen. SA; Frost. GJ: Hsie. EY; Trainer. MK; Richter. A; Burrows. JP; Peckham. SE;
Grell. GA. (2006). Satellite-observed US power plant NOx emission reductions and their impact on air
quality. Geophys Res Lett 33: L22812. http://dx.doi.org/10.1029/2006GL027749
King. GM. (1999). Characteristics and significance of atmospheric carbon monoxide consumption by soils.
Chemosphere 1: 53-63.
Kleffmann. J: Lorzer. JC: Wiesen. P; Kern. C: Trick. S: Volkamer. R; Rodenas. M; Wirtz. K. (2006).
Intercomparison of the DOAS and LOPAP techniques for the detection of nitrous acid (HONO). Atmos
Environ 40: 3640-3652.
Kleffmann. J: Wiesen. P. (2008). Technical note: Quantification of interferences of wet chemical HONO
LOPAP measurements under simulated polar conditions. Atmos Chem Phys 8: 6813-6822.
Kleindienst. TE; Hudgens. EE; Smith. DF; McElroy. FF; Bufalini. JJ. (1993). Comparison of
chemiluminescence and ultraviolet ozone monitor responses in the presence of humidity and photochemical
pollutants. Air Waste 43: 213-222.
Lam. Y; Fu. J. (2010). Corrigendum to "A novel downscaling technique for the linkage of global and regional
air quality modeling" published in Atmos. Chem. Phys., 9, 9169-9185, 2009. Atmos Chem Phys 10: 4013-
4031. http://dx.doi.org/10.5194/acp-10-4013-2010
3-232
-------
Langford. AO; Aikin. KG: Eubank. CS: Williams. EJ. (2009). Stratospheric contribution to high surface ozone
in Colorado during springtime. Geophys Res Lett 36: L12801. http://dx.doi.org/10.1029/2009gl038367
Lee. J: Kim. KH; Kim. YJ; Lee. J. (2008b). Application of a long-path differential optical absorption
spectrometer (LP-DOAS) on the measurements of NO(2), SO(2), O(3), and HNO(2) in Gwangju, Korea. J
Environ Manage 86: 750-759. http://dx.doi.Org/10.1016/i.jenvman.2006.12.044
Lefohn. AS: Shadwick. D; Oltmans. SJ. (2008). Characterizing long-term changes in surface ozone levels in
the United States (1980-2005). Atmos Environ 42: 8252-8262.
http://dx.doi.0rg/10.1016/i.atmosenv.2008.07.060
Lefohn. AS: Shadwick. D; Oltmans. SJ. (201 Ob). Characterizing changes in surface ozone levels in
metropolitan and rural areas in the United States for 19802008 and 19942008. Atmos Environ 44: 5199-
5210. http://dx.doi.0rg/10.1016/i.atmosenv.2010.08.049
Lefohn. AS: Wernli. H; Shadwick. D; Limbach. S: Oltmans. SJ: Shapiro. M. (2011). The importance of
stratospherictropospheric transport in affecting surface ozone concentrations in the western and northern
tier of the United States. Atmos Environ 45: 4845-4857. http://dx.doi.Org/10.1016/j.atmosenv.2011.06.014
Leibensperger. EM; Micklev. LJ; Jacob. DJ. (2008). Sensitivity of US air quality to mid-latitude cyclone
frequency and implications of 1980-2006 climate change. Atmos Chem Phys Discuss 8: 12253-12282.
http://dx.doi.org/10.5194/acpd-8-12253-2008
Leston. AR; Ollinson. WM; Spicer. CW: Satola. J. (2005). Potential interference bias in ozone standard
compliance monitoring. J Air Waste Manag Assoc 55: 1464-1472.
Li. Y; Lee. SR; Wu. CY. (2006c). UV-absorption-based measurements of ozone and mercury: An investigation
on their mutual interferences. Aerosol Air Qual Res 6: 418-429.
Lin. M; Fiore. AM; Horowitz. LW: Cooper. OR; Naik. V; Hollowav. J: Johnson. BJ; Middlebrook. AM;
Oltmans. SJ: Pollack. IB; Ryerson. TB; Warner. JX; Wiedinmyer. C: Wilson. J: Wyman. B. (2012).
Transport of Asian ozone pollution into surface air over the western United States in spring. J Geophys Res
117: DOOV07. http://dx.doi.org/10.1029/2011JD016961
Liu. X; Chance. K; Sioris. CE; Kurosu. TP; Spurr. RJD; Martin. RV; Fu. TM; Logan. JA; Jacob. DJ: Palmer.
PI; Newchurch. MJ; Megretskaia. IA; Chatfield. RB. (2006). First directly retrieved global distribution of
tropospheric column ozone from GOME: Comparison with the GEOS-CHEM model. J Geophys Res 111:
D02308. http://dx.doi.org/10.1029/2005JD006564
Liu. XH; Hegg. DA; Stoelinga. MT. (2001). Numerical simulation of new particle formation over the
northwest Atlantic using the MM5 mesoscale model coupled with sulfur chemistry. J Geophys Res 106:
9697-9715.
Lockwood. AL; Shepson. PB; Fiddler. MN; Alaghmand. M. (2010). Isoprene nitrates: preparation, separation,
identification, yields, and atmospheric chemistry. Atmos Chem Phys 10: 6169-6178.
http://dx.doi.org/10.5194/acp-10-6169-2010
Lu. R; Turco. RP; Jacobson. MZ. (1997). An integrated air pollution modeling system for urban and regional
scales: 1 Structure and performance. J Geophys Res 102: 6063-6079.
Luecken. DJ: Phillips. S: Sarwar. G: Jang. C. (2008). Effects of using the CB05 vs. SAPRC99 vs. CB4
chemical mechanism on model predictions: Ozone and gas-phase photochemical precursor concentrations.
Atmos Environ 42: 5805-5820. http://dx.doi.Org/10.1016/i.atmosenv.2007.08.056
Macintyre. HL; Evans. MJ. (2010). Sensitivity of a global model to the uptake of N2O5 by tropospheric
aerosol. Atmos Chem Phys 10: 7409-7414. http://dx.doi.org/10.5194/acp-10-7409-2010
Macintyre. HL; Evans. MJ. (2011). Parameterisation and impact of aerosol uptake of HO2 on a global
tropospheric model. Atmos Chem Phys 11: 10965-10974. http://dx.doi.org/10.5194/acp-ll-10965-2011
Mahajan. AS: Shaw. M; Oetjen. H; Hornsbv. KE; Carpenter. LJ: Kaleschke. L; Tian-Kunze. X; Lee. JD;
Moller. SJ: Edwards. P. (2010). Evidence of reactive iodine chemistry in the Arctic boundary layer. J
Geophys Res 115: D20303. http://dx.doi.org/10.1029/2009JD013665
3-233
-------
Martin. RV; Fiore. AM; Van Donkelaar. A. (2004). Space-based diagnosis of surface ozone sensitivity to
anthropogenic emissions. Geophys Res Lett 31: L06120. http://dx.doi.org/10.1029/2004GL019416
Maruo. YY. (2007). Measurement of ambient ozone using newly developed porous glass sensor. Sens
Actuators B 126: 485-491. http://dx.doi.Org/10.1016/i.snb.2007.03.041
Maruo. YY; Akaoka. K; Nakamura. J. (2010). Development and performance evaluation of ozone detection
paper using azo dye orange I: Effect of pH. Sens Actuators B 143: 487-493.
http://dx.doi.0rg/10.1016/i.snb.2009.09.042
Mathur. R. (2008). Estimating the impact of the 2004 Alaskan forest fires on episodic paniculate matter
pollution over the eastern United States through assimilation of satellite-derived aerosol optical depths in a
regional air quality model. J Geophys Res 113: D17302. http://dx.doi.org/10.1029/2007JD009767
McDonald-Buller. EC: Allen. DT; Brown. N; Jacob. DJ: Jaffe. D; Kolb. CE; Lefohn. AS: Oltmans. S: Parrish.
DP; Yarwood. G: Zhang. L. (2011). Establishing Policy Relevant Background (PRB) Ozone
Concentrations in the United States. Environ Sci Technol 45: 9484-9497.
http://dx.doi.org/10.1021/es2022818
McElroy. MB; Salawitch. RJ; Wofsv. SC: Logan. JA. (1986). Reductions of Antarctic ozone due to synergistic
interactions of chlorine and bromine. Nature 321: 759-762.
Milford. JB; Gao. D; Sillman. S: Blossev. P; Russell. AG. (1994). Total reactive nitrogen (NOy) as an indicator
of the sensitivity of ozone to reductions in hydrocarbon and NOx emissions. J Geophys Res 99: 3533-3542.
Millet. DB; Jacob. DJ: Boersma. KF; Fu. TM; Kurosu. TP; Chance. K; Heald. CL; Guentfaer. A. (2008). Spatial
distribution of isoprene emissions from North America derived from formaldehyde column measurements
by the OMI satellite sensor. J Geophys Res 113. http://dx.doi.org/10.1029/2007jd008950
Millet. DB; Jacob. DJ: Turquetv. S: Hudman. RC: Wu. S: Fried. A; Walega. J: Heikes. BG: Blake. PR; Singh.
HB; Anderson. BE; Clarke. AD. (2006). Formaldehyde distribution over North America: Implications for
satellite retrievals of formaldehyde columns and isoprene emission. J Geophys Res 111: D24S02.
http://dx.doi.org/10.1029/2005JD006853
Miwa. T; Maruo. YY; Akaoka. K; Kunioka. T; Nakamura. J. (2009). Development of colorimetric ozone
detection papers with high ultraviolet resistance using ultraviolet absorbers. J Air Waste Manag Assoc 59:
801-808. http://dx.doi.0rg/10.3155/1047-3289.59.7.801
Mollner. AK; Valluvadasan. S: Feng. L; Sprague. MK; Okumura. M; Milligan. DB: Bloss. WJ: Sander. SP;
Martien. PT; Harlev. RA. (2010). Rate of gas phase association of hydroxyl radical and nitrogen dioxide.
Science 330: 646-649. http://dx.doi.org/10.1126/science.1193030
Mueller. SF; Mallard. JW. (2011a). Contributions of natural emissions to ozone and PM 2.5 as simulated by the
Community Multiscale Air Quality (CMAQ) model. Environ Sci Technol 45: 4817-4823.
http://dx.doi.org/10.1021/esl03645m
Mueller. SF; Mallard. JW. (2011b). Errata in 'Contributions of natural emissions to ozone and PM 2.5 as
simulated by the Community Multiscale Air Quality (CMAQ) model' [Erratum]. Environ Sci Technol 45:
7950. http://dx.doi.org/10.1021/es2027086
Nassar. R; Logan. JA; Warden. HM; Megretskaia. IA; Bowman. KW: Osterman. GB; Thompson. AM;
Tarasick. DW: Austin. S: Claude. H; Dubev. MK; Hocking. WK; Johnson. BJ; Joseph. E; Merrill. J: Morris.
GA; Newchurch. M; Oltmans. SJ: Posny. F; Schmidlin. FJ; Vomel. H; Whiteman. DN; Witte. JC. (2008).
Validation of Tropospheric Emission Spectrometer (TES) nadir ozone profiles using ozonesonde
measurements. D15S17 (13 pp.). http://dx.doi.org/10.1029/2007jd008819
Newell. RE; Thouret. V: Cho. JYN; Stoller. P; Marenco. A; Smit. HG. (1999). Ubiquity of quasi-horizontal
layers in the troposphere. Nature 398: 316-319. http://dx.doi.org/10.1038/18642
NOAA (National Oceanic and Atmospheric Administration). (2010). The Rapid Update Cycle (RUC).
Available online at http://ruc.noaa.gov/ (accessed January 28, 2011).
3-234
-------
Nolte. CG: Gilliland. AM; Hogrefe. C: Micklev. LJ. (2008). Linking global to regional models to assess future
climate impacts on surface ozone levels in the United States. J Geophys Res 113: D14307.
http://dx.doi.org/10.1029/2007JD008497
Noziere. B; Gonzalez. NJD; Borg-karlson. AK; Pei. Y; Redebv. JP; Krejci. R; Dommen. J; Prevot. ASH;
Anthonsen. T. (2011). Atmospheric chemistry in stereo: Anew look at secondary organic aerosols from
isoprene. Geophys Res Lett 38: LI 1807. http://dx.doi.org/10.1029/2011GL047323
NFS (U.S. National Park Service). (2011). Portable Ozone Monitoring Systems (POMS). Washington, DC.
http://www.nature.nps.gov/air/studies/porto3.cfm
NRC (National Research Council). (1991). Rethinking the ozone problem in urban and regional air pollution.
Washington, DC: The National Academies Press.
NRC (National Research Council). (2007). Models in environmental regulatory decision making. Washington,
DC: National Academies Press, http://www.nap.edu/catalog/11972.html
NRC (National Research Council). (2009). Global sources of local pollution: An assessment of long-range
transport of key air pollutants to and from the United States. Washington, DC: The National Academies
Press, http://www.nap.edu/catalog.php7record id=12743
Ohira. SI; Dasgupta. PK; Schug. KA. (2009). Fiber optic sensor for simultaneous determination of atmospheric
nitrogen dioxide, ozone, and relative humidity. Anal Chem 81: 4183-4191.
http://dx.doi.org/10.1021/ac801756z
Olaguer. EP; Rappengliick. B; Lefer. B; Stutz. J: Dibb. J: Griffin. R; Brune. WH; Shauck. M; Buhr. M; Jeffries.
H; Vizuete. W: Pinto. JP. (2009). Deciphering the role of radical precursors during the Second Texas Air
Quality Study. J Air Waste Manag Assoc 59: 1258-1277. http://dx.doi.org/10.3155/1047-3289.59.ll.1258
Olszyna. KJ; Bailey. EM; Simonaitis. R: Meagher. JF. (1994). O3 and NOy relationships at a rural site. J
Geophys Res 99: 14557-14563. http://dx.doi.org/10.1029/94JD00739
Oltmans. SJ: Lefohn. AS: Harris. JM; Shadwick. DS. (2008). Background ozone levels of air entering the west
coast of the US and assessment of longer-term changes. Atmos Environ 42: 6020-6038.
http://dx.doi.0rg/10.1016/i.atmosenv.2008.03.034
Park. RJ; Stenchikov. GL; Pickering: Dickerson. RR; Allen. DJ; Kondragunta. S. (2001). Regional air pollution
and its radiative forcing: Studies with a single column chemical and radiation transport model. J Geophys
Res 106:28,751-728,770.
Parrish. DP. (2006). Critical evaluation of US on-road vehicle emission inventories. Atmos Environ 40: 2288-
2300.
Parrish. DP; Millet. DB; Goldstein. AH. (2009). Increasing ozone in marine boundary layer inflow at the west
coasts of North America and Europe. Atmos Chem Phys 9: 1303-1323. http://dx.doi.org/10.5194/acpd-8-
13847-2008
Paulot. F; Crounse. JD; Kjaergaard. HG: Kroll. JH; Seinfeld. JH; Wennberg. PP. (2009). Isoprene
photooxidation: New insights into the production of acids and organic nitrates. Atmos Chem Phys 9: 1479-
1501. http://dx.doi.org/10.5194/acp-9-1479-2009
Peeters. J: Miiller. JF. (2010). HOx radical regeneration in isoprene oxidation via peroxy radical
isomerisations. II: Experimental evidence and global impact. Phys Chem Chem Phys 12: 14227-14235.
http://dx.doi.org/10.1039/COCP00811G
Peeters. J: Nguyen. TL; Vereecken. L. (2009). HOx radical regeneration in the oxidation of isoprene. Phys
Chem Chem Phys 11: 5935. http://dx.doi.org/10.1039/B908511D
Pegues. AH; Cohan. DS: Digar. A; Douglass. C: Wilson. RS. (2012). Efficacy of recent state implementation
plans for 8-hour ozone. J Air Waste Manag Assoc 62: 252-261.
http://dx.doi.org/10.1080/10473289.2011.646049
3-235
-------
Perring. AE; Bertram. TH; Wooldridge. PJ: Fried. A; Heikes. BG: Dibb. J: Crounse. JD; Wennberg. PO; Blake.
NJ; Blake. PR; Brune. WH; Singh. HB; Cohen. RC. (2009). Airborne observations of total RONO2: New
constraints on the yield and lifetime of isoprene nitrates. Atmos Chem Phys 9: 1451-1463.
Pfister. G: Hess. PG: Emmons. LK; Lamarque. JF; Wiedinmyer. C: Edwards. DP; Petron. G: Gille. JC:
Sachese. GW. (2005). Quantifying CO emissions from the 2004 Alaskan wildfires using MOPITT CO data.
Geophys Res Lett 32: L11809.
Pinto. JP: Lefohn. AS: Shadwick. PS. (2004). Spatial variability of PM2.5 in urban areas in the United States.
J Air Waste Manag Assoc 54: 440-449.
Pokharel. SS: Bishop. GA; Stedman. DH. (2002). An on-road motor vehicle emissions inventory for Denver:
An efficient alternative to modeling. Atmos Environ 36: 5177-5184.
Pokharel. SS: Bishop. GA; Stedman. DH. (2003). Emissions reductions as a result of automobile improvement.
Environ Sci Technol 37: 5097-5101.
Pollack. AK; Lindhjem. C: Stoeckenius. TE; Iran. C: Mansell. G: Jimenez. M; Wilson. G: Coulter-Burke. S.
(2004). Final Report: Evaluation of the US EPA MOBILE6 highway vehicle emission factor model. (CRC
Project E-64). Novato, CA: ENVIRON International Corporation.
Poppe. D; Wallasch. M; Zimmermann. J. (1993). The dependence of the concentration of OH on its precursors
under moderately polluted conditions: A model study. J Atmos Chem 16: 61-78.
Rao. ST; Ku. JY; Berman. S: Zhang. K; Mao. H. (2003). Summertime characteristics of the atmospheric
boundary layer and relationships to ozone levels over the eastern United States. Pure Appl Geophys 160:
21-55.
Rappengliick. B; Dasgupta. PK; Leuchner. M; Li. Q; Luke. W. (2009). Formaldehyde and its relation to CO,
PAN, and SO2 in the Houston-Galveston airshed. Atmos Chem Phys Discuss 9: 24193-24223.
http://dx.doi.org/10.5194/acp-10-2413-2010
Rasmussen. DJ; Fiore. AM; Naik. V; Horowitz. LW: Mcginnis. SJ: Schultz. MG. (2012). Surface ozone-
temperature relationships in the eastern US: A monthly climatology for evaluating chemistry-climate
models. Atmos Environ 47: 142-153. http://dx.doi.Org/10.1016/j.atmosenv.2011.ll.021
Rastigejev. Y; Park. R; Brenner. MP; Jacob. DJ. (2010). Resolving intercontinental pollution plumes in global
models of atmospheric transport. J Geophys Res 115: D02302. http://dx.doi.org/10.1029/2009JD012568
Real. E; Law. KS: Weinzierl. B; Fiebig. M; Petzold. A; Wild. O: Methven. J: Arnold. S: Stohl. A; Huntrieser.
H; Roiger. A; Schlager. H; Stewart. D; Avery. M; Sachse. G: Browell. E; Ferrare. R; Blake. D. (2007).
Processes influencing ozone levels in Alaskan forest fire plumes during long-range transport over the North
Atlantic. J Geophys Res 112. http://dx.doi.org/10.1029/2006id007576
Reid. N; Yap. D: Bloxam. R. (2008). The potential role of background ozone on current and emerging air
issues: An overview. Air Qua! Atmos Health 1: 19-29. http://dx.doi.org/10.1007/sll869-008-0005-z
Reidmiller. PR; Fiore. AM; Jaffe. DA; Bergmann. D; Cuvelier. C: Dentener. FJ; Duncan: Bryan. N; Folberth.
G: Gauss. M; Gong. S: Hess. P: Jonson. JE; Keating. T; Lupu. A; Marnier. E; Park. R; Schultz. MG:
Shindell. DT; Szopa. S: Vivanco. MG: Wild. O: Zuber. A. (2009). The influence of foreign vs. North
American emissions on surface ozone in the US. Atmos Chem Phys 9: 5027-5042.
Reisinger. AR. (2000). Unidentified interference in DOAS measurements of ozone. Appl Spectros Rev 54: 72-
79.
Richards. NAD; Osterman. GB; Browell. EV; Hair. JW: Avery. M; Qinbin. L. (2008). Validation of
tropospheric emission spectrometer ozone profiles with aircraft observations during the intercontinental
chemical transport experiment-B. J Geophys Res 113: D16S29. http://dx.doi.org/10.1029/2007jd008815
Riediker. M; Williams. R: Devlin. R; Griggs. T; Bromberg. P. (2003). Exposure to particulate matter, volatile
organic compounds, and other air pollutants inside patrol cars. Environ Sci Technol 37: 2084-2093.
http://dx.doi.org/10.1021/es026264y
3-236
-------
Rodes. CE; Holland. DM. (1981). Variations of NO, NO2 and O3 concentrations downwind of a Los Angeles
freeway. Atmos Environ 15: 243-250.
Russell. A; Dennis. R. (2000). NARSTO critical review of photochemical models and modeling [Review].
Atmos Environ 34: 2283-2324. http://dx.doi.org/10.1016/81352-2310(99)00468-9
Rverson. TB; Buhr. MP; Frost. GJ: Goldan. PD; Hollowav. JS: Hubler. G: Jobson. BT; Kuster. WC: McKeen.
SA; Parrish. DP; Roberts. JM; Sueper. DT; Trainer. M; Williams. J: Fehsenfeld. FC. (1998). Emissions
lifetimes and ozone formation in power plant plumes. J Geophys Res 103: 22569-22583.
http://dx.doi.org/10.1029/98JD01620
Rverson. TB; Trainer. M; Hollowav. JS: Parrish. DP; Huev. LG: Sueper. DT; Frost. GJ: Donnelly. SG:
Schauffler. S: Atlas. EL; Kuster. WC: Goldan. PD; Hubler. G: Meagher. JF; Fehsenfeld. FC. (2001).
Observations of ozone formation in power plant plumes and implications for ozone control strategies.
Science 292: 719-723. http://dx.doi.org/10.1126/science.1058113
Sakugawa. H; Kaplan. IR. (1989). H2O2 and O3 in the atmosphere of Los Angeles and its vicinity: Factors
controlling their formation and their role as oxidants of SO2. J Geophys Res 94: 12957-12973.
Sarwar. G: Roselle. SJ: Mathur. R; Appel. W: Dennis. RL; Vogel. B. (2008). A comparison of CMAQ HONO
predictions with observations from the Northeast Oxidant and Particle Study. Atmos Environ 42: 5760-
5770. http://dx.doi.0rg/10.1016/i.atmosenv.2007.12.065
Schichtel. BA; Husar. RB. (2001). Eastern North American transport climatology during high- and low-ozone
days. Atmos Environ 35: 1029-1038. http://dx.doi.org/10.1016/S1352-2310(00)00370-8
Schnell. RC: Oltmans. SJ: Neely. RR; Endres. MS: Molenar. JV: White. AB. (2009). Rapid photochemical
production of ozone at high concentrations in a rural site during winter. Nat Geosci 2: 120-122.
http://dx.doi.org/10.1038/NGEO415
Seaman. NL. (2000). Meteorological modeling for air quality assessments. Atmos Environ 34: 2231-2259.
Seinfeld. JH; Pandis. SN. (1998). Atmospheric chemistry and physics: From air pollution to climate change.
New York: John Wiley & Sons.
Sexton. KG: Jeffries. HE; Jang. M; Kamens. RM; Doyle. M; Voicu. I; Jaspers. I. (2004). Photochemical
products in urban mixtures enhance inflammatory responses in lung cells. Inhal Toxicol 1: 107-114.
http://dx.doi.org/10.1080/08958370490443196
Shapiro. MA. (1980). Turbulent mixing within tropopause folds as a mechanism for the exchange of chemical
constituents between the stratosphere and troposphere. J Atmos Sci 37: 994-1004.
Sillman. S. (1995). The use of NOy, H2O2 and HNO3 as indicators for ozone-NOx-hydrocarbon sensitivity in
urban locations. J Geophys Res 100: 14175-14188.
Sillman. S: He. D; Pippin. MR; Damn. PH; Imre. DG: Kleinman. LI; Lee. JH; Weinstein-Lloyd. J. (1998).
Model correlations for ozone, reactive nitrogen, and peroxides for Nashville in comparison with
measurements: implications for O3-NOx-hydrocarbon chemistry. J Geophys Res 103: 22629-22644.
Sillman. S: He. DY. (2002). Some theoretical results concerning O3-NOx-VOC chemistry and NOx-VOC
indicators. J Geophys Res 107: 4659. http://dx.doi.org/10.1029/2001JD001123
Singh. HB; Anderson. BE; Brune. WH; Cai. C: Cohen. RC: Crawford. JH: Cubison. MJ; Czech. EP; Emmons.
L; Fuelberg. HE. (2010b). Pollution influences on atmospheric composition and chemistry at high northern
latitudes: Boreal and California forest fire emissions. Atmos Environ 44: 4553-4564.
http://dx.doi.0rg/10.1016/i.atmosenv.2010.08.026
Singh. HB; Cai. C: Kaduwela. A; Weinheimer. A; Wisthaler. A. (2012). Interactions of fire emissions and urban
pollution over California: Ozone formation and air quality simulations. Atmos Environ 56: 45-51.
http://dx.doi.0rg/10.1016/i.atmosenv.2012.03.046
Spicer. CW: Joseph. DW: Ollison. WM. (2010). A re-examination of ambient air ozone monitor interferences.
J Air Waste Manag Assoc 60: 1353-1364. http://dx.doi.org/10.3155/1047-3289.60.ll.1353
3-237
-------
Stavrakou. T; Miiller. JF; Boersma. KF; De Smedt. I; Van Per a. RJ. (2008). Assessing the distribution and
growth rates of NO x emission sources by inverting a 10-year record of NO 2 satellite columns. Geophys
Res Lett 35: L10801. http://dx.doi.org/10.1029/2008GL033521
Stedman. DH; Dabv. EE; Stuhl. F; Niki. H. (1972). Analysis of ozone and nitric oxide by a chemiluminescent
method in laboratory and atmospheric studies of photochemical smog. J Air Waste Manag Assoc 22: 260-
263.
Stevens. RK; Drago. RJ: Mamane. Y. (1993). A long path differential optical absorption spectrometer and EPA-
approved fixed-point methods intercomparison. Atmos Environ 27: 231-236.
Stutz. J; Ackermann. R; Fast. JD; Barrie. L. (2002). Atmospheric reactive chlorine and bromine at the Great
Salt Lake, Utah. Geophys Res Lett 29: 1380. http://dx.doi.org/10.1029/2002GL014812
Stutz. J; Oh. HJ; Whitlow. SI; Anderson. C: Dibb. JE; Flynn. JH; Rappengluck. B; Lefe. B. (2010).
Simultaneous DOAS and mist-chamber 1C measurements of HONO in Houston, TX. Atmos Environ 44:
40904098. http://dx.doi.0rg/10.1016/i.atmosenv.2009.02.003
Surratt. JD: Chan. AW: Eddingsaas. NC: Chan. M; Loza. CL; Kwan. AJ; Hersev. SP; Flagan. RC: Wennberg.
PO; Seinfeld. JH. (2010). Reactive intermediates revealed in secondary organic aerosol formation from
isoprene. PNAS 107: 6640-6645. http://dx.doi.org/10.1073/pnas.0911114107
Tang. Q; Prather. MJ; Hsu. J. (2011). Stratosphere-troposphere exchange ozone flux related to deep
convection. Geophys Res Lett 38: L03806. http://dx.doi.org/10.1029/2010GL046039
Tanimoto. H; Mukai. H; Hashimoto. S: Norris. JE. (2006). Intercomparison of ultraviolet photometry and gas-
phase titration techniques for ozone reference standards at ambient levels. J Geophys Res 111: Dl6313.
http://dx.doi.org/10.1029/2005JD006983
Tarasick. DW: Slater. R. (2008). Ozone in the troposphere: Measurements, climatology, budget, and trends.
Atmos Ocean 46: 93-115. http://dx.doi.org/10.3137/ao.460105
Taubman. BF; Hains. JC: Thompson. AM; Marufu. LT; Doddridge. BG: Stehr. JW: Piety. CA; Dickerson. RR.
(2006). Aircraft vertical profiles of trace gas and aerosol pollution over the mid-Atlantic United States:
Statistics and meteorological cluster analysis. J Geophys Res 111: D10S07.
http://dx.doi.org/10.1029/2005JD006196
Taubman. BF; Marufu. LT; Piety. CA; Doddridge. BG: Stehr. JW: Dickerson. RR. (2004). Airborne
characterization of the chemical, optical, and meteorological properties, and origins of a combined ozone-
haze episode over the eastern United States. J Atmos Sci 61: 1781-1793.
Thompson. AM; Stone. JB; Witte. JC: Miller. SK; Oltmans. SJ: Kucsera. TL; Ross. KL; Pickering. KE;
Merrill. JT; Forbes. G: Tarasick. DW: Joseph. E; Schmidlin. FJ; Mcmillan. WW: Warner. J: Hintsa. EJ;
Johnson. JE. (2007). Intercontinental Chemical Transport Experiment Ozonesonde Network study (IONS)
2004: 2 Tropospheric ozone budgets and variability over northeastern North America. J Geophys Res 112:
D12S13. http://dx.doi.org/10.1029/2006JD007670
Thornton. JA; Kercher. JP; Riedel. TP; Wagner. NL; Cozic. J: Hollowav. JS: Dube. WP; Wolfe. GM; Quinn.
PK; Middlebrook. AM; Alexander. B; Brown. SS. (2010). A large atomic chlorine source inferred from
mid-continental reactive nitrogen chemistry. Nature 464: 271-274. http://dx.doi.org/10.1038/nature08905
Tonnesen. GS: Dennis. RL. (2000a). Analysis of radical propagation efficiency to assess ozone sensitivity to
hydrocarbons and NO x: 1. Local indicators of instantaneous odd oxygen production sensitivity. J Geophys
Res 105: 9213. http://dx.doi.org/10.1029/1999JD900371
Tonnesen. GS: Dennis. RL. (2000b). Analysis of radical propagation efficiency to assess ozone sensitivity to
hydrocarbons and NOx: 2 Long-lived species as indicators of ozone concentration sensitivity. J Geophys
Res 105: 9227-9241. http://dx.doi.org/10.1029/1999JD900372
Tonnesen. S: Jeffries. HE. (1994). Inhibition of odd oxygen production in the carbon bond four and generic
reaction set mechanisms. Atmos Environ 28: 1339-1349. http://dx.doi.org/10.1016/1352-2310(94)90281-X
3-238
-------
Trainer. M; Parrish. DP; Buhr. MP; Norton. RB; Fehsenfeld. FC: Anlauf. KG: Bottenheim. JW: Tang. YZ;
Wiebe. HA; Roberts. JM; Tanner. RL; Newman. L; Bowersox. VC: Meagher. JF; Olszyna. KJ; Rodgers.
MO; Wang. T; Berresheim. H; Demerjian. KL; Roychowdhury. UK. (1993). Correlation of ozone with NOy
in photochemically aged air. J Geophys Res 98: 2917-2925.
Turner. AJ; Fiore. AM; Horowitz. LW: Naik. V; Bauer. M. (2012). Summertime cyclones over the Great Lakes
Storm Track from 1860–2100: variability, trends, and association with ozone pollution. Atmos
Chem Phys Discuss 12: 21679-21712. http://dx.doi.org/10.5194/acpd-12-21679-2012
U.S. Census Bureau. (2011). U.S. Census Bureau. Available online at http://www.census.gov/ (accessed
January 28, 2011).
U.S. EPA (U.S. Environmental Protection Agency). (1979b). Transfer standards for the calibration of ambient
air monitoring analyzers for ozone: Technical assistance document [EPA Report]. (EPA-600/4-79-056).
Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (1996a). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/P-93/004AF). Research Triangle Park,NC.
U.S. EPA (U.S. Environmental Protection Agency). (2006b). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/R-05/004AF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
U.S. EPA (U.S. Environmental Protection Agency). (2008a). 2005 National Emissions Inventory data and
documentation. Available online at http://www.epa.gov/ttn/chief/net/2005inventory.html (accessed
November 1, 2010).
U.S. EPA (U.S. Environmental Protection Agency). (2008c). Integrated science assessment for oxides of
nitrogen: Health criteria [EPA Report]. (EPA/600/R-08/071). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=194645
U.S. EPA (U.S. Environmental Protection Agency). (2008d). Integrated science assessment for sulfur oxides:
Health criteria [EPA Report]. (EPA/600/R-08/047F). Research Triangle Park, NC.
http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid= 198 843
U.S. EPA (U.S. Environmental Protection Agency). (2008e). National air quality: Status and trends through
2007 [EPA Report]. (EPA/454/R-08/006). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (2009d). Integrated science assessment for paniculate
matter [EPA Report]. (EPA/600/R-08/139F). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=216546
U.S. EPA (U.S. Environmental Protection Agency). (2009e). Risk and exposure assessment for review of the
secondary National Ambient Air Quality Standards for oxides of nitrogen and oxides of sulfur [EPA
Report]. (EPA/452/R-09/008A). Research Triangle Park, NC.
http://www.epa.gov/ttnnaaqs/standards/no2so2sec/cr rea.html
U.S. EPA (U.S. Environmental Protection Agency). (2010a). Air trends: Design values. Available online at
http://epa.gov/airtrends/values.html (accessed February 1, 2011).
U.S. EPA (U.S. Environmental Protection Agency). (2010b). Biogenic Emissions Inventory System (BEIS)
modeling. Available online at http://www.epa.gov/AMD/biogen.html (accessed January 28, 2011).
U.S. EPA (U.S. Environmental Protection Agency). (2010c). Integrated science assessment for carbon
monoxide [EPA Report]. (EPA/600/R-09/019F). Research Triangle Park,NC.
http ://cfpub. epa. gov/ncea/cfm/recordisplav. cfm?deid=218686
U.S. EPA (U.S. Environmental Protection Agency). (2010d). MOBILE6 vehicle emission modeling software.
Available online at http://www.epa.gov/otaq/m6.htm (accessed January 28, 2011).
U.S. EPA (U.S. Environmental Protection Agency). (2010e). Our nation's air: Status and trends through 2008
[EPA Report]. (EPA-454/R-09-002). Research Triangle Park, NC.
http ://www. epa. gov/airtrends/2010/report/fullreport.pdf
3-239
-------
U.S. EPA (U.S. Environmental Protection Agency). (201 Of). Transfer standards for calibration of air
monitoring analyzers for ozone [EPA Report]. (EPA454/B10001). Research Triangle Park, NC.
http://www.epa.gov/ttn/amtic/files/ambient/qaqc/OzoneTransferStandardGuidance.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2011 a). AirNow. Available online at
http://www.airnow.gov/ (accessed January 28, 2011).
U.S. EPA (U.S. Environmental Protection Agency). (2011e). Map monitoring sites. Available online at
http://www.epa.gov/airexplorer/monitor kml.htm (accessed January 28, 2011).
U.S. EPA (U.S. Environmental Protection Agency). (201 If). MOVES (Motor Vehicle Emission Simulator).
Available online at http://www.epa.gov/otaq/models/moves/index.htm (accessed January 28, 2011).
UNC (University of North Carolina). (2011). SMOKE (Version 2.7) [Computer Program]. Chapel Hill, NC:
University of North Carolina at Chapel Hill, Center for Environmental Modeling for Policy Development.
Retrieved from http://www.smoke-model.org/index.cfm
Univ of Leeds. NCAS. (University of Leeds, National Centre for Atmospheric Science). (2010). The master
chemical mechanism. Available online at http://mcm.leeds.ac.uk/MCM/home.htt (accessed January 28,
2011).
Utembe. SR; Hansford. GM; Sanderson. MG: Freshwater. RA; Pratt. KFE; Williams. DE; Cox. RA; Jones. RL.
(2006). An ozone monitoring instrument based on the tungsten trioxide (WO3) semiconductor. Sens
Actuators B 114: 507-512. http://dx.doi.Org/10.1016/j.snb.2005.04.049
van der Werf. GR; Randerson. JT; Giglio. L; Collatz. GJ: Kasibhatla. PS: Arellano. AF. Jr. (2006). Interannual
variability in global biomass burning emissions from 1997 to 2004. Atmos Chem Phys 6: 3423-3441.
Vardoulakis. S: Lumbreras. J: Solazzo. E. (2009). Comparative evaluation of nitrogen oxides and ozone
passive diffusion tubes for exposure studies. Atmos Environ 43: 2509-2517.
http://dx.doi.0rg/10.1016/i.atmosenv.2009.02.048
Viallon. J; Moussav. P; Norris. JE; Guenther. FR; Wielgosz. RI. (2006). A study of systematic biases and
measurement uncertainties in ozone mole fraction measurements with the NIST Standard Reference
Photometer. Metrologia 43: 441-450. http://dx.doi.Org/10.1088/0026-1394/43/5/016
Volz. A; Kiev. D. (1988). Evaluation of the Montsouris series of ozone measurements made in the nineteenth
century. Nature 332: 240-242. http://dx.doi.org/10.1038/332240aO
von Kuhlmann. R; Lawrence. MG: Poschl. U; Crutzen. PJ. (2004). Sensitivities in global scale modeling of
isoprene. Atmos Chem Phys 4: 1-17.
Wang. HQ; Jacob. DJ: Le Sager. P; Streets. DG: Park. RJ; Gilliland. AB; van Donkelaar. A. (2009a). Surface
ozone background in the United States: Canadian and Mexican pollution influences. Atmos Environ 43:
1310-1319. http://dx.doi.0rg/10.1016/i.atmosenv.2008.ll.036
Wang. J; Christopher. SA; Nair. US: Reid. JS: Prins. EM; Szykman. J: Hand. JL. (2006). Mesoscale modeling
of Central American smoke transport to the United States: 1. "Top-down" assessment of emission strength
and diurnal variation impacts. J Geophys Res 111: D05S17. http://dx.doi.org/10.1029/2005JD006416
Weaver. CP; Cooter. E; Gilliam. R; Gilliland. A; Grambsch. A; Grano. D; Hemming. B; Hunt. SW: Nolte. C:
Winner. DA; Liang. XZ; Zhu. J: Caughev. M; Kunkel. K; Lin. JT; Tao. Z; Williams. A; Wuebbles. DJ:
Adams. PJ: Dawson. JP; Amar. P; He. S: Avise. J: Chen. J: Cohen. RC: Goldstein. AH; Harlev. RA; Steiner.
AL; Tonse. S: Guenther. A; Lamarque. JF; Wiedinmyer. C: Gustafson. WI; Leung. LR; Hogrefe. C: Huang.
HC: Jacob. DJ: Micklev. LJ: Wu. S: Kinnev. PL: Lamb. B: Larkin. NK: Mckenzie. D: Liao. KJ:
Manomaiphiboon. K; Russell. AG: Tagaris. E; Lynn. BH; Mass. C: Salathe. E; O'neill. SM; Pandis. SN;
Racherla. PN; Rosenzweig. C: Woo. JH. (2009). A Preliminary Synthesis of Modeled Climate Change
Impacts on U.S. Regional Ozone Concentrations. Bull Am Meteorol Soc 90: 1843-1863.
http://dx.doi.Org/10.1175/2009BAMS2568.l
Webster. M; Nam. J: Kimura. Y; Jeffries. H; Vizuete. W: Allen. DT. (2007). The effect of variability in
industrial emissions on ozone formation in Houston, Texas. Atmos Environ 41: 9580-9593.
http://dx.doi.0rg/10.1016/i.atmosenv.2007.08.052
3-240
-------
Wells. B; Wesson. K; Jenkins. S. (2012). Analysis of recent U.S. ozone air quality data to support the O3
NAAQS review and quadratic rollback simulations to support the first draft of the risk and exposure
assessment. Wells, B; Wesson, K; Jenkins, S.
http://www.epa.gov/ttnnaaqs/standards/o3/data/20120814wells.pdf
Wernli. H; Bourqui. M. (2002). A Lagrangian "1-year climatology" of (deep) cross-tropopause exchange in the
extratropical Northern Hemisphere. J Geophys Res 107: 4021. http://dx.doi.org/10.1029/2001JD000812
Wild. O: Fiore. AM; Shindell. DT; Dohertv. RM; Collins. WJ: Dentener. FJ; Schultz. MG: Gong. S:
Mackenzie. IA; Zeng. G: Hess. P; Duncan. BN; Bergmann. DJ; Szopa. S: Jonson. JE; Keating. TJ; Zuber.
A. (2012). Modelling future changes in surface ozone: a parameterized approach. Atmos Chem Phys 12:
2037-2054. http://dx.doi.org/10.5194/acp-12-2037-2012
Williams. EJ; Fehsenfeld. FC: Jobson. BT; Kuster. WC: Goldan. PD; Stutz. J; McClenny. WA. (2006).
Comparison of ultraviolet absorbance, chemiluminescence, and DOAS instruments for ambient ozone
monitoring. Environ Sci Technol 40: 57555762. http://dx.doi.org/10.1021/es0523542
Wilson. KL; Birks. JW. (2006). Mechanism and elimination of a water vapor interference in the measurement
of ozone by UV absorbance. Environ Sci Technol 40: 6361-6367. http://dx.doi.org/10.1021/es052590c
Wimmers. AJ; Moody. JL; Browell. EV; Hair. JW: Grant. WB; Butler. CF; Fenn. MA; Schmidt. CC: Li. J:
Ridley. BA. (2003). Signatures of tropopause folding in satellite imagery. J Geophys Res 108D4: 8360.
http://dx.doi.org/10.1029/2001JD001358
Wise. EK; Comrie. AC. (2005). Meteorologically adjusted urban air quality trends in the Southwestern United
States. Atmos Environ 39: 2969-2980. http://dx.doi.Org/10.1016/j.atmosenv.2005.01.024
Warden. HM; Logan. JA; Warden. JR; Beer. R; Bowman. K; Clough. SA; Eldering. A; Fisher. BM; Gunson.
MR; Herman. RL; Kulawik. SS: Lampel. MC: Luo. M; Megretskaia. IA; Osterman. GB; Shephard. MW
(2007a). Comparisons of Tropospheric Emission Spectrometer (TES) ozone profiles to ozonesondes:
Methods and initial results. J Geophys Res 112: D03309. http://dx.doi.org/10.1029/2006JD007258
Warden. J: Liu. X; Bowman. K; Chance. K; Beer. R; Eldering. A; Gunson. M; Worden. H. (2007b). Improved
tropospheric ozone profile retrievals using OMI and TES radiances. Geophys Res Lett 34: L01809.
http://dx.doi.org/10.1029/2006GL027806
Wu. S: Micklev. LJ; Jacob. DJ: Logan. JA. (2007). Why are there large differences between models in global
budgets of tropospheric ozone? J Geophys Res 112: D05302. http://dx.doi.org/10.1029/2006JD007801
Wu. S: Micklev. LJ: Leibensperger. EM; Jacob. DJ: Rind. D; Streets. DG. (2008b). Effects of 2000-2050
global change on ozone air quality in the United States. J Geophys Res 113: D06302.
htrp://dx.doi.org/10.1029/2007JD008917
Yang. Q; Cunnold. DM; Choi. Y; Wang. Y; Nam. J: Wang. HJ; Froidevaux. L; Thompson. AM; Bhartia. PK.
(2010). A study of tropospheric ozone column enhancements over North America using satellite data and a
global chemical transport model. J Geophys Res 115: D08302. http://dx.doi.org/10.1029/2009JD012616
Yung. YL; Pinto. JP; Watson. RT; Sander. SP (1980). Atmospheric bromine and ozone perturbations in the
lower stratosphere. J Atmos Sci 37: 339-353.
Zhang. K; Wexler. A. (2008). Modeling urban and regional aerosols: Development of the UCD Aerosol
Module and implementation in CMAQ model. Atmos Environ 42: 3166-3178.
Zhang. L; Jacob. DJ: Boersma. KF; Jaffe. DA; Olson. JR; Bowman. KW: Worden. JR; Thompson. AM; Avery.
MA: Cohen. RC: Dibb. JE: Flock. FM: Fuelberg. HE: Huev. LG: McMillan. WW: Singh. HB: Weinheimer.
AJ. (2008). Transpacific transport of ozone pollution and the effect of recent Asian emission increases on
air quality in North America: An integrated analysis using satellite, aircraft, ozonesonde, and surface
observations. Atmos Chem Phys 8: 6117-6136.
Zhang. L; Jacob. DJ: Downey. NV; Wood. DA; Blewitt. D; Carouge. CC: Van donkelaar. A; Jones. DBA;
Murray. LT; Wang. Y. (2011). Improved estimate of the policy-relevant background ozone in the United
States using the GEOS-Chem global model with 1/2 2/3 horizontal resolution over North America. Atmos
Environ 45: 6769-6776. http://dx.doi.Org/10.1016/i.atmosenv.2011.07.054
3-241
-------
Zhang. L; Jacob. DJ: Logan. JA; Chance. K; Eldering. A; Bojkov. BR. (2010b). Intercomparison methods for
satellite measurements of atmospheric composition: Application to tropospheric ozone from TES and OMI.
Atmos Chem Phys 10: 4725-4739. http://dx.doi.org/10.5194/acpd-10-1417-2010
Zhang. Q; Jimenez. JL; Canagaratna. MR; Jayne. JT; Worsnop. DR. (2005). Time- and size-resolved chemical
composition of submicron particles in Pittsburgh: Implications for aerosol sources and processes. J
GeophysRes 110: 1-19.
Zhang. X; Zhuang. G: Guo. J: Yin. K; Zhang. P. (2007b). Characterization of aerosol over the Northern South
China Sea during two cruises in 2003. Atmos Environ 41: 7821-7836.
Zhang. Y. (2005). Evaluation of three probing techniques in a three-dimensional air quality model. J Geophys
Res 110: D02305. http://dx.doi.org/10.1029/2004JD005248
Ziemke. JR; Chandra. S: Duncan. BN; Froidevaux. L; Bhartia. PK; Levelt. PF; Waters. JW. (2006).
Tropospheric ozone determined from Aura OMI and MLS: Evaluation of measurements and comparison
with the Global Modeling Initiative's Chemical Transport Model. J Geophys Res 111: D19303.
http://dx.doi.org/10.1029/2006JD007089
Zimmermann. J: Poppe. D. (1993). Nonlinear chemical couplings in the tropospheric NOx-HOx gas phase
chemistry. JAtmos Chem 17: 141-155.
3-242
-------
4 EXPOSURE TO AMBIENT OZONE
4.1 Introduction
The 2006 O3 AQCD (U.S. EPA. 2006b) evaluated O3 concentrations and exposures
in multiple microenvironments, discussed methods for estimating personal and
population exposure via monitoring and modeling, analyzed relationships between
personal exposure and ambient concentrations, and discussed the implications of
using ambient O3 concentrations as an estimate of exposure in epidemiologic studies.
This chapter presents new information regarding exposure to ambient O3 which
builds upon the body of evidence presented in the 2006 O3 AQCD. A brief summary
of findings from the 2006 O3 AQCD is presented at the beginning of each section as
appropriate.
Section 4.2 presents general exposure concepts describing the relationship between
ambient pollutant concentrations and personal exposure. Section 4.3 describes
exposure measurement techniques and studies that measured personal, ambient,
indoor, and outdoor concentrations of O3 and related pollutants. Section 4.4 presents
material on parameters relevant to exposure estimation, including activity patterns,
averting behavior, and population proximity to ambient monitors. Section 4.5
describes techniques for modeling local O3 concentrations, air exchange rates,
microenvironmental concentrations, and personal and population exposure.
Section 4.6 discusses the implications of using ambient O3 concentrations to estimate
exposure in epidemiologic studies, including several factors that contribute to
exposure error.
4.2 General Exposure Concepts
A theoretical model of personal exposure is presented to highlight measurable
quantities and the uncertainties that exist in this framework. An individual's
time-integrated total exposure to O3 can be described based on a
compartmentalization of the person's activities throughout a given time period:
-S"
Equation 4-1
where ET = total exposure over a time -period of interest, Cj = airborne O3
concentration at microenvironmentj, and dt = portion of the time-period spent in
microenvironment / Equation 4-1 can be decomposed into a model that accounts for
exposure to O3, of ambient (Ea) and nonambient (Ena) origin of the form:
4-1
-------
ET — EO. + Ena
Equation 4-2
Ambient O3 is formed through photochemical reactions involving NOX, VOCs, and
other compounds, as described in Chapter 3.. Although nonambient sources of O3
exist, such as O3 generators and laser printers, these sources are specific to
individuals and may not be important sources of population exposure. Ozone
concentrations generated by ambient and nonambient sources are subject to spatial
and temporal variability that can affect estimates of exposure and influence
epidemiologic effect estimates. Exposure parameters affecting interpretation of
epidemiologic studies are discussed in Section 4.6.
This assessment focuses on the ambient component of exposure because this is more
relevant to the NAAQS review. Assuming steady-state outdoor conditions, Ea can be
expressed in terms of the fraction of time spent in various outdoor and indoor
microenvironments (U.S. EPA. 2006c: Wilson et al. 2000):
— / i J o^-o ' / i Ji
Equation 4-3
where/= fraction of the relevant time period (equivalent to dt in Equation 4-1).
subscript o = index of outdoor microenvironments, subscript i = index of indoor
microenvironments, subscript o,i = index of outdoor microenvironments adjacent to a
given indoor microenvironment /', and Finf , = infiltration factor for indoor
microenvironment i. Equation 4-3 is subject to the constraint Ł/"<, + Ł/j = 1 to reflect
the total exposure over a specified time period, and each term on the right hand side
of the equation has a summation because it reflects various microenvironmental
exposures. Here, "indoors" refers to being inside any aspect of the built environment,
e.g., home, office buildings, enclosed vehicles (automobiles, trains, buses), and/or
recreational facilities (movie theaters, restaurants, bars). "Outdoor" exposure can
occur in parks or yards, on sidewalks, and on bicycles or motorcycles. Assuming
steady state ventilation conditions, the infiltration factor is a function of the
penetration (P) of O3 into the microenvironment, the air exchange rate (a) of the
microenvironment, and the rate of O3 loss (k) in the microenvironment;
In epidemiologic studies, the central-site ambient concentration, Ca, is often used in
lieu of outdoor microenvironmental data to represent these exposures based on the
availability of data. Thus it is often assumed that C0 = Ca and that the fraction of
time spent outdoors can be expressed cumulatively as/0; the indoor terms still retain
a summation because infiltration differs among different microenvironments. If an
epidemiologic study employs only Ca, then the assumed model of an individual's
4-2
-------
exposure to ambient O3, first given in Equation 4-3, is re-expressed solely as a
function of Ca:
Equation 4-4
The spatial variability of outdoor O3 concentrations due to meteorology, topography,
varying precursor emissions and O3 formation rates; the design of the epidemiologic
study; and other factors determine whether or not Equation 4-4 is a reasonable
approximation for Equation 4-3. These equations also assume steady-state
microenvironmental concentrations. Errors and uncertainties inherent in use of
Equation 4-4 in lieu of Equation 4-3 are described in Section 4.6 with respect to
implications for interpreting epidemiologic studies. Epidemiologic studies often use
concentration measured at a central site monitor to represent ambient concentration;
thus a, the ratio between personal exposure to ambient O3 and the ambient
concentration of O3, is defined as:
Equation 4-5
Combination of Equation 4-4 and Equation 4-5 yields:
« = /o +
Equation 4-6
where a varies between 0 and 1 . If a person's exposure occurs in a single
microenvironment, the ambient component of a microenvironmental O3
concentration can be represented as the product of the ambient concentration and
.Fjnf. The U.S. EPA (2006c) noted that time-activity data and corresponding estimates
of Finf for each microenvironmental exposure are needed to compute an individual's
a with accuracy. In epidemiologic studies, a is assumed to be constant in lieu of
time-activity data and estimates of Finf, which can vary with building and
meteorology-related air exchange characteristics. If important local outdoor sources
and sinks exist that are not captured by central site monitors, then the ambient
component of the local outdoor concentration may be estimated using dispersion
models, land use regression models, receptor models, fine scale CTMs or some
combination of these techniques. These techniques are described in Section 4.5.
4-3
-------
4.3 Exposure Measurement
This section describes techniques that have been used to measure
microenvironmental concentrations of O3 and personal O3 exposures as well as
results of studies using those techniques. Previous studies from the 2006 O3 AQCD
are described along with newer studies that evaluate indoor-outdoor concentration
relationships, associations between personal exposure and ambient monitor
concentration, and multipollutant exposure to other pollutants in conjunction with
O3. Tables are provided to summarize important study results.
4.3.1 Personal Monitoring Techniques
As described in the 2006 O3 AQCD, passive samplers have been developed and
deployed to measure personal exposure to O3 (Grosjean and Hisham, 1992; Kanno
and Yanagisawa, 1992). Widely used versions of these samplers utilize a filter coated
with nitrite, which is converted to nitrate by O3 and then quantified by a technique
such as ion chromatography (Koutrakis et al., 1993). This method has been licensed
and marketed by Ogawa, Inc., Japan (Ogawa & Co, 2007). The cumulative sampling
and the detection limit of the passive badges makes them mainly suitable for
monitoring periods of 24 hours or greater, which limits their ability to measure short-
term daily fluctuations in personal O3 exposure. Longer sampling periods give lower
detection limits; use of the badges for a 6-day sampling period yields a detection
limit of 1 ppb, while a 24-hour sampling period gives a detection limit of
approximately 5-10 ppb. This can result in a substantial fraction of daily samples
being below the detection limit (Sarnat et al., 2006a; Sarnat et al., 2005), which is a
limitation of past and current exposure studies. Development of improved passive
samplers capable of shorter-duration monitoring with lower detection limits would
enable more precise characterization of personal exposure in multiple
microenvironments with relatively low participant burden.
The nitrite-nitrate conversion reaction has also been used as the basis for an active
sampler consisting of a nitrite-coated glass tube through which air is drawn by a
pump operating at 65 mL/min (Geyh et al., 1999; Geyh et al., 1997). The reported
detection limit is 10 ppb-h, enabling the quantification of O3 concentrations
measured over a few hours rather than a full day (Geyh et al., 1999).
A portable active O3 monitor based on the UV photometric technique used for
stationary monitors (Chapter 3J has recently been approved as a FEM (75 FR 22126).
This monitor includes a Nafion tube in the inlet line to equilibrate humidity, reducing
the effect of humidity changes in different microenvironments (Wilson and Birks.
2006). Its size and weight (approximately 10^20^30 cm; 2 kg) make it suitable for
use in a backpack configuration. The monitors are currently used by the U.S.
National Park service as stationary monitors to measure O3 in several national parks
(Chapter 3_). Future improvements and continued miniaturization of real-time O3
4-4
-------
monitors can yield highly time-resolved personal measurements to further evaluate
O3 exposures in specific situations, such as near roadways or while in transit.
4.3.2 Indoor-Outdoor Concentration Relationships
Several studies summarized in the 2006 O3 AQCD, along with some newer studies,
have evaluated the relationship between indoor O3 concentration and the O3
concentration immediately outside the indoor microenvironment. These studies show
that the indoor concentration is often substantially lower than the outdoor
concentration unless indoor sources are present. Low indoor O3 concentrations can
be explained by reactions of O3 with surfaces and airborne constituents. Studies have
shown that O3 is deposited onto indoor surfaces where reactions produce secondary
pollutants such as formaldehyde (Reiss et al., 1995b; Reiss et al., 1995a). However,
the indoor-outdoor relationship is greatly affected by the air exchange rate; under
conditions of high air exchange rate, such as open windows, the indoor O3
concentration may approach the outdoor concentration. Thus, in rooms with open
windows, the indoor-outdoor (I/O) ratio may approach 1.0. Table 4-1 summarizes
I/O ratios and correlations reported by older and more recent studies, with discussion
of individual studies in the subsequent text. In general, I/O ratios range from about
0.1 to 0.4, with some evidence for higher ratios during the O3 season when
concentrations are higher.
Ozone concentrations near and below the monitor detection limit cause uncertainty in
I/O ratios, because small changes in low concentration values cause substantial
variation in resulting ratios. This problem is particularly acute in the non-ozone
season when ambient O3 concentrations are low. Further improvements in
characterization of microenvironmental O3 concentrations and I/O ratios will rely on
improved monitoring. Until new monitoring techniques are available and can be used
in the field, past studies summarized in the 2006 O3 AQCD remain relevant to
consider along with more recent studies in evaluating the relationship between indoor
and outdoor O3 concentrations.
4-5
-------
Table 4-1 Relationships between indoor and outdoor O3 concentration.
Sample
Study Location Years/ Season Population duration Ratio3 Correlation
June -Sept 1995 0 „.
and May 1996
Upland,
California
Gevh et al. June - Sept 1995 _.... _ . „ __ .._
?5000) and May 1996 Chlldren 6 days °'36 NR
Mountain
Communities,
Southern
California
Oct!999956-Apr
Upland & _ .. . ,
Mountain Entlre Perlod
Micro-
environ- Concentration/
ment Comment Detection limit (ppb)
Home Mean (SD) Indoor
11.8(9.2)
Outdoor
Air-conditioned 482(122)
Ratio: Indoor mean/
outdoor mean lndoor
3.2 (3.9)
Outdoor
21.1 (10.7)
Indoor
21.4(14.8)
Outdoor
Window ventilation 601(171)
Ratio: indoor mean/
outdoor mean Indoor
2.8 (4.2)
Outdoor
35.7 (9.3)
LODM.O
Fraction above LOD
Indoor 80.3%
Outdoor99.95%
4-6
-------
Study
Avol et al.
(1998a)
Romieu et
al.(1998a)
Lee et al.
(2004a)
Sample
Location Years/ Season Population duration Ratio3 Correlation
0.37
SD: 0.25
Feb-Dec1994 IQR. 0.58
0.07-
Southern 0.45
California Nr' ^" ....
Summer (late 0.43
June - late Sept) SD: 0.29
0.32
Non-summer NR
SD:0.21
0.20
SD:0.18
Mexico City, Sept 1993 - July 7 or 14 days °^b NR
Mexico 1994 Rgnge.
0.01-
1.00
0 1
Nashville, TN Summer 1994 Children 1 week NR
SD; 0.18
Micro-
environ- Concentration/
ment Comment Detection limit (ppb)
Mean (SD)
Indoor
13(12)
Outdoor
Home Ratio: each pair of 3?(19)
nome values -31 (ia/
LOD:5
Mean (SD)
Indoor
7-day: 5 (4)
14-day: 7 (5)
H Ratio: each pair of Outdoor
values 7-day: 27 (10)
14-day: 37 (12)
LOD:
7-day: 2.4-2.9
14-day: 1.2-3.5
Indoor
Range of Weekly
Means: 1.6-3.1
Fraction above LOD,
Range:
Ratio: Indoor mean/ 14-87%
ome outdoor mean Outdoor
Range of Weekly
Means: 18.6-25.9
Fraction above LOD:
100%
LOD: 1.2
4-7
-------
Sample
Study Location Years/Season Population duration
Ratio3
Correlation
Micro-
environ-
ment Comment
Concentration/
Detection limit (ppb)
Satchewa,
Canada
Summer 2007
All age
groups
5 days
0.13
NR
Home
Ratio: Indoor mean/
outdoor mean
Mean (SD)
Indoor
0.7 (0.8)
Outdoor
5.4(1.3)
LOD: NR
Fraction above field
LOD: Indoor: <50%
Outdoor: NR
Liu et al.
(1995)
Toronto,
Canada
Winter 1992
Summer 1992
Summer 1992
1 week
0.07
SD:0.10
0.40
SD: 0.29
Ratio: each pair of
values
All age
groups
NR
Home
12h
Summer 1992
0.30
SD: 0.32
0.43
SD: 0.54
Daytime
Ratio: each pair of
values
Nighttime
Ratio: each pair of
values
Mean (SD)
Indoor
1.6(4.1)
Outdoor
15.4(6)
LOD: 1.05
Indoor. NR
Outdoor. NR
Indoor
7.1 (12.6)
Outdoor
19.1 (10.8)
LOD: 14.7
Indoor
6.2 (9.5)
Outdoor
9.4(10.2)
LOD: NR
4-8
-------
Sample
Study Location Years/ Season Population duration Ratio3 Correlation
Chi,dren ^ 0,5
Romieu et Mexico City, Sept 1993 - July .|R
al.(1998a) Mexico 1994 Nhc
'SS ><%• 0.30-
S3 "*" ""
Blondeau et La Rochelle, _ . „„„„ -.... „ , .5 "!e- . ._
al. (2005) France Spring 2000 Children 2 weeks 0.00- NR
U.4o
Micro-
environ- Concentration/
ment Comment Detection limit (ppb)
Mean (SD)
Indoor
Ratio: each pair of 6 (2.8)
values LOD:0.7-1.3
Outdoor
41 (8.2)
Indoor
School 5-day: 22 (16.1)
10-day: 22(16.0)
LOD:
rsecdiate,y outside 5^^
10 day: 0.3-1.6
Outdoor
5-day: 73 (21 .5)
10-day: 56 (17.9)
Range*
Indoor
0-57
No air conditioning Outdoor
School Ratio: Indoor mean/ rj-68
outdoor mean . „.-.. .
'Estimated from
Figure 1 , showing
two of eight schools
4-9
-------
Sample
Study Location Years/ Season Population duration Ratio3 Correlation
July 2009 0.10
Loeez; Prague, Dec 2009 All age 0.30
Aparicio et Czech y 1 month NR
aU2011) Republic groups
July 2009 -Mar
2010 Overall
0.51
Riediker et North A ~ . onm Aj )t 0 . MD
al. (2003) Carolina ""
Micro-
environ- Concentration/
ment Comment Detection limit (ppb)
Mean*
Indoor. 2
Outdoor. 18.2
Mean*
No heating or air Indoor. 1 .3
Historic conditioning Outdoor. 4.4
Library Ratio: Indoor mean/ 'Estimated from Fig.
outdoor mean 2
Range
Indoor: 1-2.5
Outdoor: 4.1 -21 .9
LOD: 0.5
Mean (SD)
In-vehicle
Vehicle Ratio: Indoor mean/ 11.7(15.9)
outdoor mean Roadside
22.8(13.3)
LOD: 10
aMean value unless otherwise indicated
"Median
LOD = limit of detection; NR = not reported; SD = standard deviation
4-10
-------
Geyh et al. (2000) measured 6-day indoor and outdoor concentrations at 116 homes
in southern California, approximately equally divided between the community of
Upland and several mountain communities. The extended sampling period resulted in
a relatively low detection limit (1 ppb) for the passive samplers used and a large
fraction of samples above the detection limit; over 80% of the indoor samples and
virtually all of the outdoor samples were above the detection limit. The Upland
homes were nearly all air-conditioned, while the mountain community homes were
ventilated by opening windows. During the O3 season, the indoor O3 concentration
averaged over all homes was approximately 24% of the overall mean outdoor
concentration in Upland (11.8 versus 48.2 ppb), while in the mountain communities,
the indoor concentration was 36% of the outdoor concentration (21.4 versus
60.1 ppb). This is consistent with the increased air exchange rate expected in homes
using window ventilation. In the non-ozone season, when homes are likely to be
more tightly closed to conserve heat, the ratios of indoor to outdoor concentration
were 0.15 (3.2 versus 21.1 ppb) and 0.08 (2.8 versus 35.7 ppb) in Upland and the
mountain communities, respectively. Avol et al. (1998a) observed a mean I/O ratio
of 0.37 for 239 matched 24-hour samples collected between February and December
at homes in the Los Angeles area. The I/O ratio during summer was somewhat higher
than the non-summer I/O ratio (0.43 versus 0.32). The authors also reported a
correlation of 0.58 between the 24-h avg indoor concentration and the outdoor
concentration, which was only slightly higher than the correlation between the indoor
concentration and the concentration at the neighborhood fixed-site monitor (0.49).
Substantially higher summer I/O ratios were reported in a study in Toronto, Canada
(Liu et al.. 1995). which found summer I/O ratios of 0.30-0.43, in comparison with a
winter I/O ratio of 0.07. Romieu et al. (1998a) reported a mean I/O ratio of 0.20 in
145 homes in Mexico City, Mexico, for 7- or 14-day cumulative samples, with the
highest ratios observed in homes where windows were usually open during the day
and where there was no carpeting or air filters. Studies conducted in Nashville, TN,
and Regina, Saskatchewan (Canada) reported mean residential I/O ratios of
approximately 0.1 (Heroux et al., 2010; Lee et al., 2004a).
Investigators have also measured I/O ratios for non-residential microenvironments,
including schools and vehicles. Romieu et al. (1998a) reported that O3 concentrations
measured during school hours (10-day cumulative sample, 5 h/day) were 30-40% of
concentrations immediately outside the schools, while overall I/O ratios (14-day
cumulative sample, 24 h/day) were approximately 15%. The authors attribute this
discrepancy to increased air exchange during the school day due to opening doors
and windows. Air exchange was also identified as an important factor in the I/O
ratios measured at eight French schools (Blondeau et al.. 2005). In this study, the I/O
ratios based on simultaneous continuous measurements ranged from 0-0.45,
increasing with decreasing building tightness. A historical library building in Prague,
Czech Republic with no heating or air conditioning (i.e., natural ventilation) was
observed to have ratios of one-month indoor and outdoor concentrations ranging
from 0.10-0.30 during a nine-month sampling campaign, with the highest ratios
reported in Nov-Dec 2009 and the lowest ratios during July-Aug 2009 (Lopez-
Aparicio et al., 2011). Indoor concentrations were relatively constant (approximately
3-7 |_ig/m3 or 2-3 ppb), while outdoor concentrations were lower in the winter
4-11
-------
(9-10 j^g/ m3 or about 5 ppb) than in the summer (35-45 \igl m3 or about 20 ppb).
This seasonal variation in outdoor concentrations coupled with homogeneous indoor
concentrations, together with increased wintertime air exchange rate due to higher
indoor-outdoor temperature differences, is likely responsible for the observed
seasonal pattern in I/O ratios.
Exposures in near-road, on-road and in-vehicle microenvironments are likely to be
more variable and lower in magnitude than those in other microenvironments due to
reaction of O3 with NO and other combustion emissions. Depending on wind
direction, O3 concentrations near the roadway have been found to be 20-80% of
ambient concentrations at sites 400 meters or more distant from roads
(Section 3.6.2.1). A study on patrol cars during trooper work shifts reported in-
vehicle 9-h concentrations that were approximately 51% of simultaneously measured
roadside concentrations (mean of 11.7 versus 22.8 ppb) (Riediker et al., 2003).
4.3.3 Personal-Ambient Concentration Relationships
Several factors influence the relationship between personal O3 exposure and ambient
concentration. Due to the lack of indoor O3 sources, along with reduction of ambient
O3 that penetrates into enclosed microenvironments, indoor and in-vehicle O3
concentrations are highly dependent on air exchange rate and therefore vary widely
in different microenvironments. Ambient O3 varies spatially due to reactions with
other atmospheric species, especially near busy roadways where O3 concentrations
are decreased by reaction with NO (Section 3.6.2.1). This is in contrast with
pollutants such as CO and NOX, which show appreciably higher concentrations near
the roadway than several hundred meters away (Karner et al.. 2010). Ozone also
varies temporally over multiple scales, with generally increasing concentrations
during the daytime hours, and higher O3 concentrations during summer than in
winter. An example of this variability is shown in Figure 4-1, taken from a personal
exposure study conducted by Chang et al. (2000).
In this figure, hourly personal exposures are seen to vary from a few ppb in some
indoor microenvironments to tens of ppb in vehicle and outdoor microenvironments.
The increase in ambient O3 concentration during the day is apparent from the
outdoor monitoring data. In comparison, ambient PM2.5 exhibits less temporal
variability over the day than O3, although personal exposure to PM2.5 also varies by
microenvironment. This combined spatial and temporal variability for O3 results in
varying relationships between personal exposure and ambient concentration.
Correlations between personal exposure to O3 and corresponding ambient
concentrations, summarized in Table 4-2, exhibit a wide range (generally 0.3-0.8,
although both higher and lower values have been reported), with higher correlations
generally observed in outdoor microenvironments, high building ventilation
conditions, and during the summer season. Low O3 concentrations indoors and
during the winter lead to a high proportion of personal exposures below the sampler
detection limit, which may partially explain the low correlations observed in some
4-12
-------
studies under those conditions. Studies report varying correlations over a range of
averaging times, with no clear trend. Ratios of personal exposure to ambient
concentration, summarized in Table 4-3, are generally lower in magnitude (typically
0.1-0.3), and are also variable, with increasing time spent outdoors associated with
higher ratios. The next two subsections describe studies that have reported personal-
ambient correlations and slopes for a variety of seasons, locations, and populations.
Personal and Outdoor PM2 s and O3:
Baltimore, MD, August 12, 1998
-0- Outdoor I'M-,
• 0 - Outdoor O,
. .77 . IVrsonal O,
I
90 -
80 -
70-
50 -
3 40 -
30 -
20 -
10 -
0 -
60
- 40
- 30
- 20
- 10
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
walking kitchen study/ room health walking food car ma!! mall restaurant car
TVuwtti to ruoin club own
Clock Hour (EST)
Note: the notation below each clock hour shows the location or activity during that hour.
Source: Reprinted with permission of Air and Waste Management Association (Chang et al.. 2000).
Figure 4-1 Variation in hourly personal and ambient concentrations of O3 and
PM2.5 in various microenvironments during daytime hours.
Ozone concentrations near and below the passive sampler detection limit lead to
uncertainty in personal-ambient correlations and ratios. Correlations are reduced in
magnitude by values below the detection limit because noise obscures the underlying
signal in the data, while ratios tend to fluctuate widely at low concentration since
small changes in measured values cause large relative changes in resulting ratios.
4-13
-------
As with I/O ratios, this problem is particularly acute in the non-ozone season when
ambient O3 concentrations are low. Improved characterization of the relationship
between personal exposure and ambient concentration will depend on the use of
recent improved monitoring techniques to accurately capture low O3 concentrations,
preferably at high time resolution to facilitate evaluation of the effect of activity
pattern on exposure (Section 4.3.1). While data from studies using new monitoring
techniques become available, past studies summarized in the 2006 O3 AQCD (U.S.
EPA. 2006b) remain relevant to consider along with more recent studies in
evaluating personal-ambient concentration relationships.
Personal-Ambient Correlations. Correlations between personal exposure and
ambient O3 concentrations have been evaluated in several research studies, many of
which were conducted prior to 2005 and are discussed in the 2006 O3 AQCD. Some
studies evaluated subject-specific, or longitudinal correlations, which describe
multiple daily measurements for a single individual. These studies indicate the
inter-individual variability of personal-ambient correlations. Another type of
correlation is a pooled correlation, which combines data from multiple individuals
over multiple monitoring periods (e.g., days), providing an overall indicator of the
personal-ambient relationship for all study subjects. A third type of correlation is a
community-average correlation, which correlates average exposure across all study
subjects with fixed-site monitor concentrations. Community-average correlations are
particularly informative for interpreting time-series epidemiologic studies, in which
ambient concentrations are used as a surrogate for community-average exposure.
However, few studies report this metric; this represents another opportunity for
improvement of future personal exposure studies. Table 4-2 summarizes studies
reporting personal-ambient correlations, and the studies in the table are discussed in
the subsequent text.
The results of these studies generally indicate that personal exposures are moderately
well correlated with ambient concentrations, and that the ratio of personal exposure
to ambient concentration is higher in outdoor micro environments and during the
summer season. In some situations, a low correlation was observed, and this may be
due in part to a high proportion of personal measurements below the detection limit
of the personal sampler [e.g., Sarnat et al. (2000)]. Apart from this, correlations do
not appear to be strongly dependent on concentration. The effect of season is unclear,
with mixed evidence on whether higher correlations are observed during the O3
season. Chang et al. (2000) measured hourly personal exposures in multiple
microenvironments and found that the pooled correlation between personal exposure
and ambient concentration was highest for outdoor microenvironments
(r = 0.68-0.91). In-vehicle microenvironments showed moderate to high correlations
(0.57-0.72). Correlations in residential indoor microenvironments were very low
(r = 0.05 - 0.09), with moderate correlations (0.34-0.46) in other indoor
microenvironments such as restaurants and shopping malls. Liard et al. (1999)
evaluated community-average correlations based on 4-day mean personal O3
exposure measurements for adults and children and found a relatively high
correlation (r = 0.83) with ambient concentrations, even with only 18-69% of the
personal measurements above the detection limit. Sarnat et al. (2000) studied a
4-14
-------
population of older adults in Baltimore, MD, and found that longitudinal correlations
between 24-h personal exposure and ambient concentration varied by subject and
season, with somewhat higher correlations observed in this study during summer
(mean = 0.20) than in winter (mean = 0.06). Although the fraction of samples above
the detection limit was not reported separately for the older adults in this study, in the
larger study of which this population is a part, only a few percent of samples were
above the detection limit, with less than 1% above the detection limit during the
winter (see Table 4-3) (Koutrakis et al. 2005). This may account for the low
observed correlations, particularly in winter. Some evidence was presented that
subjects living in well-ventilated indoor environments have higher correlations than
those living in poorly ventilated indoor environments, although exceptions to this
were also observed. Ramirez-Aguilar et al. (2008) measured 48- to 72-h personal
exposures of four groups of asthmatic children aged 6-14 in Mexico City, Mexico,
during 1998-2000. A moderate pooled correlation (r = 0.35) was observed between
these exposures and corresponding ambient concentrations.
4-15
-------
Table 4-2
Study
Chang et al.
(2000)
Liard et al.
(1999)
Correlations between personal and ambient O3 concentration.
Sample
Location Years/Season Population duration Correlation Study Type
Summer 1998 0.91
Winter 1999 0.77
Summer 1998 0.68
Winter 1999 0.86
Summer 1998 0.72
Winter 1999 0.57
Baltimore, MD Older adults 1 h Pooled
Summer 1998 0.09
Winter 1999 0.05
Summer 1998 0.34
Winter 1999 0.46
Paris, France Summer 1996 All age groups 4 day 0.83 Community-
averaged
Concentration
Comment (ppb)
Outdoor near Summer1998
roadway Personal
Median (Range)'
10.0 (-11. 3-76)
Outdoor away v '
from road Mean (SD): 15.0
(18.3)
Ambient*
In vehicle Range: 12.2-59.8
Winter 1999
Indoors- Personal
residence Median (Range):
05 (-03-1 2 2)
Mean (SD): 1.1
(1.7)
, , ,, Ambient"
Indoors-other
Range: 12.2-24.6
'Estimated from
Figure 3
Range
Children
Persona/*
-0.1-3.9
Ambient*
14.8-30.7
Adults
Persona/*
0.9-2.7
Ambient*
19.8-27.5
'Estimated from
Figure 3
Personal
Detection
limit (ppb)
Summer 1998
LOD:17.2
Winter 1999
LOD:12.0
Fraction
above LOD:
NR
LOD: 1.5 ppb
Fraction
above LOD:
Children: 15-
69%
Adults: 18-
34%
4-16
-------
Sample Concentration
Study Location Years/Season Population duration Correlation Study Type Comment (ppb)
Mean (SD)
°'20 Personal
Summer SD: °'28 3.5 (3.0)
95% Cl: 0.06, . ,. ,
Q34 Ambient
POOO) 6t a' Baltimore, MD Older adults 24 h Longitudinal 37.3(8.3)
0.06 Personal
Winter SD: °'34 ° (1 '8)
95% Cl: Ambient
-0.88,0.24 17.8(10.3)
Mean (SD)
Persona/
Ambient
23(12)
Range
Persona/*
Hearth clinic 24 h „ „ poo|ed 0-25% of time „„„,
Ambient*
5.3-30.1
Persona/*
Brauerand Vancouver, Summers 1992 Ł,^L,nre 24 h 0.42 Pooled l'^™M°* al~13'8
Brook (1997) Canada and 1993 counselors time outdoors Ambient*
11.2-26.1
Persona/*
3.2-43
Farmworkers 6-1 4 h (work poo|ed 100% of time Ambient*
shift) outdoors g g_4g Q
'Estimated from
Figure 2
Personal
Detection
limit (ppb)
LOD: 6.6
Fraction
above LOD:
NR
LOD: 5.5
Fraction
above LOD:
NR
LOD: NR
Fraction
above LOD:
NR
LOD
24h: 17 ppb
12h: 12 ppb
Fraction
above LOD:
NR
4-17
-------
Study Location
Ramirez- ... ...
T — ~. T Mexico City,
(2008) MexluJ
Delfino et al. San Diego,
(1 996) California
Sample Concentration
Years/Season Population duration Correlation Study Type Comment (ppb)
Mean (Range)
Persona/
April 2000 children
Ambient
33.3(12.5-64.6)
Mean (SD)
Persona/
0 45
September and Asthmatic 12-h ' 11.6(11.2)
October 1993 children daytime Range: 0.35- °° e Ambient
12h:43(17)
1hr max: 68 (30)
Personal
Detection
limit (ppb)
LOD: NR
Fraction
above LOD:
NR
LOD: 8.67
Fraction
above LOD:
53%
LOD = limit of detection; NR = not reported; SD = standard deviation
4-18
-------
Consistent with hourly microenvironment-specific results from the Chang et al.
(2000) study described above, studies have found moderate to high personal-ambient
correlations for individuals spending time outdoors. A moderate pooled correlation
of 0.61 was reported between 24-h avg personal and central-site measurements by
Linn et al. (1996) for a population of southern California schoolchildren who spent
an average of 101-136 minutes per day outdoors. The authors also report a
correlation of 0.70 between central-site measurements and concentrations outside the
children's schools. Although the average school outdoor concentration (34 ppb) was
higher than the average central-site concentration (23 ppb) and the average personal
exposure concentration was lower (5 ppb) than the central-site value, the similarity
between the correlations in this study indicate that central-site monitor concentrations
can represent personal exposures in addition to representing local outdoor
concentrations. A study in Vancouver, BC provided another illustration of the effect
of outdoor microenvironments on personal-ambient relationships by comparing three
groups spending different amounts of time outdoors: health clinic workers (0-25% of
sampling period outdoors), camp counselors (7.5-45% of sampling period outdoors),
and farm workers (100% of sampling period outdoors) (Brauer and Brook. 1997).
Health clinic workers and camp counselors were monitored 24 h/day, while farm
workers were monitored during their work shift (6-14 hours). In this study, the
pooled correlations between personal exposure and fixed-site concentration were not
substantially different among the groups (r = 0.60, 0.42, and 0.64, respectively), even
though the farm workers experienced the highest concentrations. The ratios of
personal exposure to fixed-site monitor concentration increased among the groups
with increasing amount of time spent outdoors (0.35, 0.53, and 0.96, respectively).
This indicates that temporal variations in personal exposure to O3 are driven by
variations in ambient concentration, even for individuals that spend little time
outdoors.
Personal-Ambient Ratios. Studies indicate that the ratio between personal O3
exposure and ambient concentration varies widely, depending on activity patterns,
housing characteristics, and season. Higher personal-ambient ratios are generally
observed with increasing time spent outside, higher air exchange rate, and in seasons
other than winter. Table 4-3 summarizes the results of several such studies discussed
in the 2006 O3 AQCD together with newer studies showing the same pattern of
results.
O'Neill et al. (2003) studied a population of shoe cleaners working outdoors in
Mexico City, Mexico, and presented a regression model indicating a 0.56 ppb
increase in 6-h personal exposure for each 1 ppb increase in ambient concentration.
Regression analyses by Sarnat et al. (2005) and Sarnat et al. (2001) for 24-hour data
from mixed populations of children and older adults in Baltimore, MD (Sarnat et al.,
2001), and Boston, MA (Sarnat et al., 2005), found differing results between the two
cities, with Baltimore subjects showing a near-zero slope (0.01) during the
summertime while Boston subjects showed a positive slope of 0.27 ppb personal
exposure per 1 ppb ambient concentration. In both cities, the winter slope was near
zero. The low slope observed in Baltimore may have been due to differences in time
spent outdoors, residential ventilation conditions, or other factors. The intercept in
4-19
-------
Baltimore was more than half of the median personal exposure (1.84 versus 2.5 ppb),
and only 6.6% of the personal samples were above the detection limit, suggesting
that noise in the data may also have contributed to the low observed regression slope.
However, other studies with a relatively large fraction of personal values below the
detection limit reported slopes of 0.27 or above (Sarnat et al., 2006a; Brauer and
Brook. 1997). Xue et al. (2005) measured 6-day personal exposure of children in
southern California and found that the average ratio of personal exposure to ambient
concentration was relatively stable throughout the year at 0.3. These authors also
regressed personal exposures on ambient concentration after adjusting for time-
activity patterns and housing characteristics and found a slope of 0.54 ppb/ppb, with
the regression R2 value of 0.58. Unadjusted regression slopes were not presented.
It should also be noted that the ratio and slope would not be expected to be identical
unless the intercept and other regression parameters were effectively zero.
A few additional studies have been published since the 2006 O3 AQCD comparing
personal exposures with ambient concentrations, and these findings generally
confirm the conclusions of the 2006 O3 AQCD that ventilation conditions, activity
pattern, and season may impact personal-ambient ratios. Sarnat et al. (2006a)
measured 24-hour personal exposures for a panel of older adults in Steubenville, OH
during summer and fall 2000. Subjects were classified as high-ventilation or low-
ventilation based on whether they spent time in indoor environments with open
windows. Regression of personal exposures on ambient concentration found a higher
slope for high-ventilation subjects compared with low-ventilation subjects in both
summer (0.18 versus 0.08) and fall (0.27 versus 0.20). Suh and Zanobetti (2010)
reported an average 24-hour personal exposure of 2.5 ppb as compared to 24-hour
ambient concentration of 29 ppb for a group of individuals with either recent MI or
diagnosed COPD in Atlanta. A similar result was observed in Detroit, where the
mean 24-hour personal exposure across 137 participants in summer and winter was
2.1 ppb, while the mean ambient concentration on sampling days was 25 ppb
(Williams et al.. 2009b). Although no personal exposures were measured, McConnell
et al. (2006) found that average 24-hour home outdoor O3 concentrations were within
6 ppb of O3 concentrations measured at central-site monitors in each of three
southern California communities, with a combined average home outdoor
concentration of 33 ppb compared to the central-site average of 36 ppb. In Mexico
City, Mexico, Ramirez-Aguilar et al. (2008) regressed 48- to 72-hour personal
exposures of four groups of asthmatic children aged 6-14 with ambient
concentrations and found slope of 0.17 ppb/ppb after adjustment for distance to the
fixed-site monitor, time spent outdoors, an interaction term combining these two
variables, and an interaction term representing neighborhood and study group.
4-20
-------
Table 4-3 Ratios of personal to ambient O3 concentration.
Years/ Sample
Study Location Season Population duration Ratio3 Slope
Inter-
cept Study Type
Concentration/
Detection limit
Comment (ppb)
Summer
1998
Older adults
and children
..R
„ „.
1.84
t-value:1.21
Sarnat et al.
(2001)
Koutrakis et
al. (2005)
Median (IQR)
Personal*
2.5 (0-6.4)
LOD: 6.6
Fraction above LOD:
6.6%
Ambient*
36 (31 -43)
Baltimore, MD
24 h
Longitudinal
Older adults,
0.46
t-value: 0.03
COPD
Personal*
1.1 (-0.6-1.9)
LOD: 5.5
Fraction above LOD:
0.2%
Ambient*
18 (8.6-26)
'Estimated from
Figure 1
4-21
-------
Study
Location
Years/
Season
Population
Sample
duration
Ratio3
Slope
Inter-
cept
Study Type Comment
Concentration/
Detection limit
(PPb)
Range
Personal*
»*
5.3-30.1
LOD: 17
Vancouver,
?n™mer!
1992 and
1993
Camf
counselors
24 h
0.53
NR
NR
Pooled
75-45%
of time
.. __
outdoors
Farmworkers
0.96
NR
NR
Pooled
Personal*
0.1-13.8
Ambient*
11.2-26.1
LOD: 17
Personal*
3.2-43
LOD: 12
'Estimated from
Figure 2
O'Neill etal.
(2003)
Mexico CitV' AP1996U'y Shoe Cleaners
0.40
6 h 0.37b
SD:0.22
0 56
95% Cl: NR Longitudinal
0.43-0.69
Mean (SD)
Persona/
34.4 (22.3)
Ambient
84.0 (24.8)
LOD: 21.1 (20.6)
4-22
-------
Study
Location
Years/
Season
Population
Sample
duration
Ratio3
Slope
Inter-
cept
Study Type Comment
Concentration/
Detection limit
(ppb)
Summer
NR
0.27
95% Cl:
0.18-0.37
NR
Sarnat et al.
(2005)
24 h
Lonflitudinal
Winter
NR
0.04
95% CI'
0 00 0 07
NR
Range of Means
Personal
4.8-7.6
LOD: 7.0
Fraction above LOD:
23.8%
Ambient
Mean (SD): 22.7-
Range of Means
Personal
0.1-2.5
LOD: 4.9
Fraction above LOD:
3.2%
Ambient
14.0-21.8
Xue et al.
(2005)
Southern
California
JMne1gg|"
ay
Children
6 day
„ _
'
SD. 0.13
NR
NR
Longitudinal
O3 season (May-
Sept)
Persona/*
Median (IQR):
22(14-30)
Ambient*
Median (IQR):
53 (44-67)
Non-03 season
(Oct-Apr)
Personal"
Median (IQR):
6(5-10)
Ambient*
Median (IQR):
26(14-32)
LOD: NR
'Estimated from
Figure 2
4-23
-------
Years/ Sample Inter-
Study Location Season Population duration Ratio3 Slope cept Study Type
Comment
Concentration/
Detection limit
(PPb)
Summer
Sarnat et al.
(2006a)
Steuben-ville,
OH
Older adults
24 h
Fall
NR
NR
NR
NR
NR
NR
0.15
SE:0.02
t-value:
7.21
R2: 0.24
0.18
SE:0.03
t-value:
7.34
R2: 0.27
0.08
SE:0.04
t-value:
1.89
R2:0.19
0.27
SE: 0.03
t-value:
8.64
R2: 0.25
0.27
SE:0.04
t-value:
7.38
R2: 0.33
0.20
SE:0.05
t-value:
3.90
R2:0.12
NR
NR
NR
NR
NR
NR
All individuals
Mean (SD)
Personal 5.3 (5.2)
Fraction above LOD:
8.2%
Ambient 29.3 (13.4)
Fraction above LOD:
94%
LOD: 12.7
High-
ventilation
Low-
ventilation
Longitudinal
All individuals
Mean (SD)
Personal 3.9 (4.4)
Fraction above LOD:
8.4%
Ambient 16.0(8.1)
Fraction above LOD:
71%
LOD: 10.7
High-
ventilation
Low-
ventilation
4-24
-------
Study Location
3 ass*
Years/ Sample Inter-
Season Population duration Ratio3 Slope cept Study Type
0 17
SE- 002
Dec.1998- Asthmatic 48 h to 95% Cl:
Apr. 2000 children 72 h 0.13-0.21 Pooled
p-value:
0.00
Concentration/
Detection limit
Comment (ppb)
Mean (Range)
Persona/
7.8 (0.2-30.9)
Ambient
33.3(12.5-64.6)
LOD: NR
B Mean value unless otherwise indicated
b Median
IQR = interquartile range; LOD = limit of detection; NR = not reported; SD = standard deviation
4-25
-------
4.3.4 Co-exposure to Other Pollutants and Environmental Stressors
Exposure to ambient O3 occurs in conjunction with exposure to a complex mixture of
ambient pollutants that varies over space and time. Multipollutant exposure is an
important consideration in evaluating health effects of O3 since these other pollutants
have either known or potential health effects that may impact health outcomes due to
O3. The co-occurrence of high O3 concentrations with high heat and humidity may
also contribute to health effects. This section presents data on relationships between
overall personal O3 exposure and exposure to other ambient pollutants, as well as
co-exposure relationships for near-road O3 exposure.
4.3.4.1 Personal Exposure to Ozone and Copollutants
Personal exposure to O3 shows variable correlation with personal exposure to other
pollutants, with differences in correlation depending on factors such as instrument
detection limit, season, city-specific characteristics, time scale, and spatial variability
of the copollutant. Suh and Zanobetti (2010) reported Spearman rank correlation
coefficients during spring and fall between 24-h avg O3 measurements and
copollutants of 0.14, 0.00, and -0.03 for PM2.5, EC, andNO2, respectively. Titration
of O3 near roadways is likely to contribute to the low or slightly negative correlations
with the traffic-related pollutants EC and NO2. The somewhat higher correlation
with PM2.s may reflect the influence of air exchange rate and time spent outdoors on
co-exposures to ambient PM2 5 and O3. Overall, the copollutant correlations are quite
small, which may be due to the very low personal exposures observed in this study
(2-3 ppb), likely to be near or below the detection limit of the passive sampler over a
24-hour period. Chang et al. (2000) measured hourly personal exposures to PM2 5
and O3 in summer and winter in Baltimore, Maryland. Correlations between PM2 5
and O3 were 0.05 and -0.28 in summer and winter, respectively. Results indicate
personal O3 exposures were not significantly associated with personal PM2 5
exposures in either summer or winter. These non-significant correlations may be
attributed in part to the relatively low personal O3 exposures observed in this study;
in both summer and winter, the mean personal O3 exposure was below the calculated
limit of detection.
Studies conducted in Baltimore, MD (Sarnat et al.. 2001). and Boston, MA (Sarnat et
al.. 2005). found differing results for the correlation between 24-h avg personal O3
and personal PM2 5 exposures, particularly during the winter season. Sarnat et al.
(2001) found a positive slope when regressing personal exposures of both total PM2 5
(0.21) and PM25 of ambient origin (0.22) against personal O3 exposures during the
summer season, but negative slopes (-0.05 and -0.18, respectively) during the winter
season. The summertime slope for personal PM2 5 exposure versus personal O3
exposure was much higher for children (0.37) than for adults (0.07), which may be
4-26
-------
the result of different activity patterns. This team of researchers also found a positive,
although higher, summer slope between 24-h avg personal O3 and personal PM2.5 in
Boston (0.72) (Sarnat et al, 2005). However, the winter slope was positive (1.25)
rather than negative, as in Baltimore. In both cities during both seasons, there was a
wide range of subject-specific correlations between personal O3 and personal PM2.5
exposures, with some subjects showing relatively strong positive correlations (>0.75)
and others showing strong negative correlations (<-0.50). The median correlation in
both cities was slightly positive in the summer and near zero (Boston) or slightly
negative (Baltimore) in the winter. These results indicate the potential effects of city-
specific characteristics, such as housing stock and building ventilation patterns, on
relationships between O3 and copollutants.
The lack of long-term exposure assessment studies limits evaluation of long-term
correlations between O3 exposure and copollutant exposure. Although some long-
term epidemiologic studies have reported copollutant correlations for fixed-site
monitor concentrations or city-wide averages used as exposure metrics, no clear
pattern is apparent. Correlations with PM concentrations range from less than 0.2 to
nearly 0.9. For example, the long-term correlation between 30-yr mean (1973-1992)
PMio and the 30-yr mean of 8-h avg (9 am to 5 pm) O3 was 0.88 for participants in
the Adventist Health Smog study (AHSMOG) (McDonnell et al., 1999a). Jerrett et
al. (2009) reported a moderate correlation of 0.56 between two-year average O3 and
PM2.s in 86 U.S. metropolitan areas. In the Southern California Children's Health
Study, the correlation between 1994-2000 average O3 and PM2 5 was 0.33 for 1-h
daily max O3, but only 0.18 for the 1994-2000 mean of 8-h avg (10 am - 6 pm) O3
concentrations (Gauderman et al., 2004). Similar correlations were reported in this
study between these O3 metrics and PMi0. For children participating in the National
Health Interview Survey and living in U.S. metropolitan areas, the correlation
between 2000-2004 O3 concentrations and PM2 5 and PMio concentrations was 0.29
and 0.55, respectively (Akinbami et al.. 2010). For NO2, near-zero or negative
correlations have been reported, consistent with atmospheric chemistry involving
NO2 and O3. Correlations with NO2 in the Children's Health Study were 0.10 and -
0.11 for 1994-2000 mean 1-h daily max O3 and 8-h avg O3, respectively
(Gauderman et al.. 2004). This is similar to the value of -0.05 reported by Akinbami
et al. (2010). However, the AHSMOG study reported a correlation of 0.61 between
30-yr averages of O3 and NO2, possibly reflecting similar overall levels of air
pollution experienced by the participants (McDonnell et al.. 1999a). This lack of a
consistent pattern makes it difficult to draw conclusions regarding long-term
correlations between O3 and copollutants.
To the extent that short-term concentrations drive long-term patterns, some insight
may be provided by an analysis of short-term correlations between O3 and other
criteria pollutants, such as is provided in Section 3.6.4. Warm-season 8-h daily max
O3 concentrations are generally positively correlated with co-located 24-h avg
measurements of other criteria pollutants (Figure 3-57). Median correlations range
from approximately 0.15 to 0.55 for CO, SO2, NO2, PMio, and PM2.5, in that order.
In contrast, year-round 8-h daily max O3 data show negative median correlations
with CO and NO2, positive correlations with PMio and PM2 5, and essentially zero
4-27
-------
correlation with SO2. This reflects mostly negative correlations between O3 and all
pollutants during wintertime, as shown in Figure 3-56. Titration of O3 near roadways
also likely contributes to overall negative correlations with NO2 and CO. Positive
correlations between O3 and PM2.5 during the summertime can be partly explained
by meteorological conditions favoring increased formation of both secondary PM
and O3. Strong positive correlations can influence the interpretation of epidemiologic
results, potentially complicating the ability to identify the independent effect of a
pollutant.
4.3.4.2 Near-Road Exposure to Ozone and Copollutants
Beckerman et al. (2008) measured both 1-week and continuous concentrations of O3,
NO, NO2, NOX, PM2.5, PMi.o, and several VOCs (the BTEX compounds, MTBE,
hexane, and THC) in the vicinity of heavily traveled (annual average daily traffic
[AADT] >340,000) roadways in Toronto, Canada. Passive samplers were deployed
for one week in August 2004. Ozone concentrations were negatively correlated with
all pollutants, with the exception of VOCs at one of the monitoring sites which were
suspected of being influenced by small area sources. Site specific correlations are
given in Figure 4-2. Correlations were -0.77 to -0.85 for NO2, -0.48 to -0.62 for NO,
and -0.55 to -0.63 for NOX. Pooled correlations using data from both sites were
somewhat lower in magnitude. PM2 5 and PMi 0 correlations were -0.35 to -0.78 and
-0.34 to -0.58, respectively. At the monitoring site not influenced by small area
sources, O3-VOC correlations ranged from -0.41 to -0.66.
Beckerman et al. (2008) also made on-road measurements of multiple pollutants with
a instrumented vehicle. Concentrations were not reported, but correlations between
O3 and other pollutants were negative and somewhat greater in magnitude (i.e., more
negative) than the near-road correlations. SO2, CO, and BC were measured in the
mobile laboratory, although not at the roadside, and they all showed negative
correlations with O3 when the data were controlled for site. Correlations for
continuous concentrations between O3 and copollutants were somewhat lower than
the 1-week correlations, except for O3-PM25 correlations. Correlations were -0.90,
-0.66, -0.77, and -0.89 for NO2, NO, NOX, and PMLO respectively. The continuous
O3-PM2 5 correlation was -0.62, which is in the range of the 1-week correlation.
4-28
-------
-0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1
Pearson Correlation Coefficient
Source: Data from: Beckerman et al. (2008)
Figure 4-2 Correlations between 1-week concentrations of O$ and
copollutants measured near roadways.
4.3.4.3 Indoor Exposure to Ozone and Copollutants
Ambient O3 that infiltrates indoors reacts with organic compounds and other
chemicals to form oxidized products, as described in Section 3.2.3.1 as well as the
2006 O3 AQCD. It is anticipated that individuals are exposed to these reaction
products, although no evidence was identified regarding personal exposures.
The reactions are similar to those occurring in the ambient air, as summarized in
Chapter 3_. For example, O3 can react with terpenes and other compounds from
cleaning products, air fresheners, and wood products both in the gas phase and on
surfaces to form particulate and gaseous species, such as formaldehyde (Chen et al..
2011: Shu and Morrison. 2011: Aoki and Tanabe. 2007: Reiss et al.. 1995b). Ozone
has also been shown to react with material trapped on HVAC filters and generate
airborne products (Beko et al.. 2007: Hvttinen et al.. 2006). Potential oxygenated
reaction products have been found to act as irritants (Anderson et al.. 2007).
indicating that these reaction products may have health effects separate from those of
O3 itself (Weschler and Shields. 1997). Ozone may also react to form other oxidants,
which then go on to participate in additional reactions. White et al. (2010) found
evidence that HONO, or other oxidants, may have been present during experiments
to estimate indoor OH concentrations; indicating complex indoor oxidant chemistry.
Rates of these reactions are dependent on indoor O3 concentration, temperature, and
air exchange rate, making estimation of exposures to reaction products difficult.
4-29
-------
4.4 Exposure-Related Metrics
In this section, parameters are discussed that are relevant to the estimation of
exposure, but are not themselves direct measures of exposure. Time-location-activity
patterns, including behavioral changes to avoid exposure, have a substantial
influence on exposure and dose. Proximity of populations to ambient monitors may
influence how well their exposure is represented by measurements at the monitors,
although factors other than distance play an important role as well.
4.4.1 Activity Patterns
The activity pattern of individuals is an important determinant of their exposure.
Variation in O3 concentrations among various microenvironments means that the
amount of time spent in each location, as well as the level of activity, will influence
an individual's exposure to ambient O3. The effect of activity pattern on exposure is
explicitly accounted for in Equation 4-3 by the fraction of time spent in different
mi croenvironments.
Activity patterns vary both among and within individuals, resulting in corresponding
variations in exposure across a population and over time. Large-scale human activity
databases, such as those developed for the National Human Activity Pattern Survey
(NHAPS) (Klepeis et al.. 2001) or the Consolidated Human Activity Database
(CHAD) (McCurdy et al.. 2000). which includes NHAPS data together with other
activity study results, have been designed to characterize exposure patterns among
much larger population subsets than can be examined during individual panel studies.
The complex human activity patterns across the population (all ages) are illustrated
in Figure 4-3 (Klepeis et al.. 2001). which is presented to illustrate the diversity of
daily activities among the entire population as well as the proportion of time spent in
each microenvironment. For example, about 25% of the individuals reported being
outdoors or in a vehicle between 2:00 and 3:00 p.m., when daily O3 levels are
peaking, although about half of this time was spent in or near a vehicle, where O3
concentrations are likely to be lower than ambient concentrations.
Time spent in different locations has also been found to vary by age. Table 4-4
summarizes NHAPS data reported for four age groups, termed Very Young (0-4
years), School Age (5-17 years), Working (18-64 years), and Retired (65+ years)
(Klepeis et al.. 1996). The working population spent the least time outdoors, while
the school age population spent the most time outdoors. NHAPS respondents aged 65
and over spent somewhat more time outdoors than adults aged 18-64, with a greater
fraction of time spent outdoors at a residence. Children aged 0-4 also spent most of
their outdoor time in a residential outdoor location. On average, the fraction of time
spent outdoors by school age respondents was 2.62 percentage points higher than
working respondents, corresponding to approximately 38 minutes more time
outdoors per day. Although not accounting for activity level, this increased time
4-30
-------
spent outdoors is consistent with evidence in Chapter Ł suggesting that younger and
older age groups are more at risk for O3-related health effects.
Table 4-4 Mean fraction of time spent in outdoor locations by various age
groups in the NHAPS study.
Age Group
0-4 yr
5-1 7 yr
18-64yr
65+ yr
Source: Data from Kleoeis et al.
Residential-Outdoor
5.38%
5.05%
2.93%
4.48%
(1996).
Other Outdoor
0.96%
2.83%
2.33%
1 .27%
Total Outdoors
6.34%
7.88%
5.26%
5.75%
Together with location, exertion level is an important determinant of exposure.
Table 4-5 summarizes ventilation rates for different age groups at several levels of
activity as presented in Table 6-2 of the EPA's Exposure Factors Handbook (U.S.
EPA. 201 Ib). Most of the age-related variability is seen for moderate and high
intensity activities, except for individuals under 1 year. For moderate intensity,
ventilation rate increases with age through childhood and adulthood until age 61,
after which a moderate decrease is observed. Ventilation rate is most variable for
high intensity activities. Children aged 1 to <11 years have ventilation rates of
approximately 40 L/min, while children aged 11+ and adults have ventilation rates of
approximately 50 L/min. The peak is observed for the 51 to <61 age group, at 53
L/min, with lower ventilation rates for older adults.
A dramatic increase in ventilation rate occurs as exercise intensity increases. For
children and adults <31 years, high intensity activities result in nearly double the
ventilation rate for moderate activity, which itself is nearly double the rate for light
activity. Children have other important differences in ventilation compared to adults.
As discussed in Chapter 5_, children tend to have a greater oral breathing contribution
than adults, and they breathe at higher minute ventilations relative to their lung
volumes. Both of these factors tend to increase dose normalized to lung surface area.
4-31
-------
Table 4-5 Mean ventilation rates (L/min) at different activity levels for
different age groups.
Age Group
Birth to <1 yr
1 to <2 yr
2 to <3 yr
3 to <6 yr
6 to <1 1 yr
11 to <16yr
16 to <21 yr
21 to <31 yr
31 to <41 yr
41 to <51 yr
51 to <61 yr
61 to <71 yr
71 to <81 yr
81 + yr
Sleep or Nap
3.0
4.5
4.6
4.3
4.5
5.0
4.9
4.3
4.6
5.0
5.2
5.2
5.3
5.2
Sedentary/Passive
3.1
4.7
4.8
4.5
4.8
5.4
5.3
4.2
4.3
4.8
5.0
4.9
5.0
4.9
Light Intensity
7.6
12
12
11
11
13
12
12
12
13
13
12
12
12
Moderate
Intensity
14
21
21
21
22
25
26
26
27
28
29
26
25
25
High Intensity
26
38
39
37
42
49
49
50
49
52
53
47
47
48
Source: Data from Exposure Factors Handbook (U.S. EPA. 2011b).
Longitudinal activity pattern information is also an important determinant of
exposure, as different people may exhibit different patterns of time spent outdoors
over time due to age, sex, employment, and lifestyle-dependent factors. These
differences may manifest as higher mean exposures or more frequent high-exposure
episodes for some individuals. The extent to which longitudinal variability in
individuals contributes to the population variability in activity and location can be
quantified by the ratio of between-person variance to total variance in time spent in
different locations and activities (the intraclass correlation coefficient, ICC). Xue et
al. (2004) quantified ICC values in time-activity data collected by Harvard
University for 160 children aged 7-12 years in Southern California (Geyh et al.,
2000). For time spent outdoors, the ICC was approximately 0.15, indicating that 15%
of the variance in outdoor time was due to between-person differences. The ICC
value might be different for other population groups.
4-32
-------
100
ooooooooooooooooooooooooo
ppppppppppppppppppppppppp
ol i— CN rl TT v"i vc r~- 00 G*i O '— r-j i—i r-j f^i "^f >/"l ^C I"-- CO <3* O •—i f>]
Time of Day
Source: Reprinted with permission of Nature Publishing Group (Klepeis et al., 2001).
Figure 4-3 Distribution of time that NHAPS respondents spent in ten
microenvironments based on smoothed 1-min diary data.
The EPA's National Exposure Research Laboratory (NERL) has consolidated many
of the most important human activity databases into one comprehensive database
called the Consolidated Human Activity Database (CHAD). The current version of
CHAD contains data from nineteen human activity pattern studies (including
NHAPS), which were evaluated to obtain over 33,000 person-days of 24-hour human
activities in CHAD (McCurdy et al., 2000). The surveys include probability-based
recall studies conducted by EPA and the California Air Resources Board, as well as
real-time diary studies conducted in individual U.S. metropolitan areas using both
probability-based and volunteer subject panels. All ages of both sexes are represented
in CHAD. The data for each subject consist of one or more days of sequential
activities, in which each activity is defined by start time, duration, activity type, and
microenvironment classification (i.e., location). Activities vary from one minute to
one hour in duration, with longer activities being subdivided into clock-hour
durations to facilitate exposure modeling. CHAD also provides information on the
level of exertion associated with each activity, which can be used by exposure
models, including the APEX model (Section 4.5.3). to estimate ventilation rate and
pollutant dose.
4-33
-------
4.4.2 Ozone-Averting Behavior
Individuals can reduce their exposure to O3 by altering their behaviors, such as by
staying indoors, scheduling outdoor activity during periods of low O3 concentration,
and by reducing activity levels or time spent being active outdoors on high-O3 days.
To assist the public in avoiding exposure to air pollution on days with high pollutant
concentrations, EPA has developed an information tool known as the Air Quality
Index (AQI) to provide information to the public on ambient levels of pollutants and
the potential for individuals to experience health effects (U.S. EPA, 201 la). The AQI
describes the potential for health effects from O3 (and other individual pollutants) in
six color-coded categories of air-quality, ranging from good (green), moderate
(yellow), unhealthy for sensitive groups (orange), unhealthy (red), very unhealthy
(purple), and hazardous (maroon). The levels are associated with descriptors of the
likelihood of health effects and the populations most likely to be affected. For
example, the orange level indicates that the general population is not likely to be at
risk, but susceptible groups may experience health effects. These advisories
explicitly state that children, older adults, people with lung disease, and those who
are active outdoors may be at greater risk from exposure to air pollution. Forecasted
and actual conditions typically are reported to the public during high-O3 months
through local media outlets, using various versions of this air-quality categorization
scheme. People are advised to change their behavior to reduce exposure depending
on predicted O3 concentrations and the likelihood of risk. Behavioral
recommendations include moving outdoor activities to times when air quality is
better, and reducing activity levels or the time spent being active outdoors on high-
O3 days. Staying indoors to reduce exposure is only recommended when the AQI is
at or above the very unhealthy range.
Evidence of individual averting behaviors in response to advisories has been found in
several studies, especially for potentially susceptible populations, such as children,
older adults, and asthmatics. Reduced time spent outdoors was reported in an activity
diary study in 35 U.S. cities (Mansfield et al., 2006), which found that asthmatic
children who spent at least some time outdoors reduced their total time spent
outdoors by an average of 30 min on a code red O3 day relative to a code green,
yellow, or orange day; however, the authors noted that there was appreciable
variation in both the overall amount of time spent outdoors and the reduction in
outdoor time on high O3 days among asthmatic children. Bresnahan et al. (1997)
examined survey data collected during 1985-86 from a panel of adults in the Los
Angeles area, many of whom had compromised respiratory function, by an averting
behavior model. A regression analysis indicated that individuals with smog-related
symptoms spent about 12 minutes less time outdoors over a two-day period for each
10 ppb increase in O3 concentration above 120 ppb. Considering that the average
daily maximum O3 concentration at the time was approximately 180 ppb on days
when the then-current standard (1-h max of 120 ppb) was exceeded, this implies that
those individuals spent about 40 minutes less time outside per day on a typical high
O3 day compared to days with O3 concentrations below the standard. However, the
behavior was not specifically linked to exceedances or air quality alerts.
4-34
-------
The fraction of individuals who reduce time spent outdoors, or restrict their
children's outdoor activity, has been found to vary based on health status. In the
Bresnahan et al. (1997) study, 40 percent of respondents reported staying indoors on
days when air quality was poor. Individuals who reported experiencing smog-related
symptoms were more likely to take the averting actions, although the presence of
asthma or other chronic respiratory conditions did not have a statistically significant
effect on behavior. A study of parents of asthmatic children flVIcDermott et al.. 2006)
suggests that parents are aware of the hazard of outdoor air pollution and the official
alerts designed to protect them and their children. It also suggests that a majority of
parents (55%) comply with recommendations of the alerts to restrict children's
outdoor activity, with more parents of asthmatics reporting awareness and
responsiveness to alerts. However, only 7% of all parents complied with more than
one-third of the advisories issued (McDermott et al.. 2006). Wen et al. (2009)
analyzed data from the 2005 Behavioral Risk Factor Surveillance System (BRFSS)
and indicated that people with asthma are about twice as likely as people without
asthma to reduce their outdoor activities based on either media alerts of poor air
quality (31% versus 16%) or individual perception of air quality (26% versus 12%).
Respondents who had received advice from a health professional to reduce outdoor
activity when air quality is poor were more likely to report a reduction based on
media alerts, both for those with and without asthma. In a study of randomly selected
individuals in Houston, TX and Portland, OR, Semenza et al. (2008) found that a
relatively small fraction of survey respondents (9.7% in Houston, 10.5% in Portland)
changed their behaviors during poor air quality episodes. This fraction is appreciably
lower than the fraction reported for people with asthma in the Wen et al. (2009)
study, although it is similar to the fraction reported in that study for those without
asthma. Most of the people in the Semenza et al. (2008) study reported that their
behavioral changes were motivated by self-perception of poor air quality rather than
an air quality advisory. It should be noted that the McDermott et al. (2006), Wen et
al. (2009), and Semenza et al. (2008) studies evaluated air quality in general and
therefore are not necessarily specific to O3.
Commuting behavior does not seem to change based on air quality alerts. A study in
the Atlanta area showed that advisories can raise awareness among commuters but do
not necessarily result in a change in an individual's travel behavior (Henry and
Gordon. 2003). This finding is consistent with a survey for 1,000 commuters in
Denver, Colorado, which showed that the majority (76%) of commuters heard and
understood the air quality advisories, but did not alter their commuting behavior
(Blanken et al.. 2001).
Some evidence is available for other behavioral changes in response to air quality
alerts. Approximately 40 percent of the respondents in the Los Angeles study by
Bresnahan et al. (1997) limited or rearranged leisure activities, and 20 percent
increased use of air conditioners. As with changes in time spent outdoors, individuals
who reported experiencing smog-related symptoms, but not those with asthma or
chronic respiratory conditions, were more likely to take the averting actions. Other
factors influencing behavioral changes, such as increased likelihood of averting
behavior among high school graduates, are also reported in the study. In a separate
4-35
-------
Southern California study, attendance at two outdoor facilities (i.e., a zoo and an
observatory) was reduced by 6-13% on days when smog alerts were announced, with
greater decreases observed among children and older adults (Neidell. 2010, 2009).
The studies discussed in this section indicate that averting behavior is dependent on
several factors, including health status and lifestage. People with asthma and those
experiencing smog-related symptoms reduce their time spent outdoors and are more
likely to change their behavior than those without respiratory conditions. Children
and older adults appear more likely to change their behavior than the general
population. Commuters, even when aware of air quality advisories, tend not to
change their commuting behavior.
4.4.3 Population Proximity to Fixed-Site Ozone Monitors
The distribution of O3 monitors across urban areas varies between cities
(Section 3.6.2.1). and the population living near each monitor varies as well.
Monitoring sites in rural areas are generally located in national or state parks and
forests, and these monitors may be relevant for exposures of exercising visitors as
well as those who live in similar locations. They also serve as an important source of
data for evaluating ecological effects of O3 (Chapter 9). Rural monitors tend to be
less affected than urban monitors by strong and highly variable anthropogenic
sources of species participating in the formation and destruction of O3 (e.g., onroad
mobile sources) and more highly influenced by regional transport of O3 or O3
precursors (Section 3.6.2.2). This may contribute to less diel variability in O3
concentration than is observed in urban areas.
A variety of factors determine the siting of the O3 monitors that are part of the
SLAMS network reporting to AQS. As discussed in Section 3.5.6. the number and
location of required O3 monitors in an urban area depend on O3 concentration and
population, among other factors. Areas classified as serious, severe, or extreme
nonattainment have additional monitoring requirements. Generally, high-O3 urban
areas with a population of 50,000 or greater are required to have at least one monitor;
in low- or moderate-concentration areas, the minimum population for a required
monitor is 350,000. Most large U.S. cities have several monitors, as shown in
Figure 3-76 through Figure 3-95.
As an illustration of the location of O3 monitors and their concentrations with respect
to population density, Figure 4-4 through Figure 4-6 present this information for
Atlanta, Boston, and Los Angeles, the three cities selected for detailed analysis in
Chapter 3_. They represent a cross-section with respect to geographic distribution, O3
concentration, layout, geographic features, and other factors. The maps show the
location of O3 monitors, identified by the same letters as in Chapter 3_ to facilitate
intercomparisons, along with the 2007-2009 mean 8-h daily max O3 concentration
for perspective on the variation in O3 concentration across the urban area. Population
density at the census block group level is also presented on the maps.
4-36
-------
I
100 Kilometers
| I I I | I I T
0510 20 Kilometers
Monitor Mean 8-h Daily
Max Ozone, 2007-2009
22 - 44 ppb
45 - 49 ppb
50 - 59 ppb
60 - 69 ppb
70 - 79 ppb
Interstate Highways
Major Highways
Atlanta CSA Block Groups
2009 Population per Sq Km
0-565
566- 1275
| 1276-2173
^H 2174-3600
j^H 3601 - 6247
^H 6248- 13320
Figure 4-4 Map of the Atlanta CSA including Os monitor locations and major
roadways with respect to census block group population density
estimates for 2009.
4-37
-------
I "• I ' I ' ' I
0 30 60 120 Kilometers
Monitor Mean 8-h Daily
Max Ozone, 2007-2009
Boston CSA Block Groups
2009 Population per Sq Km
22 - 44 ppb
45 - 49 ppb
50 - 59 ppb
60 - 69 ppb
70 - 79 ppb
Interstate Highways
Major Highways
Figure 4-5 Map of the Boston CSA including O3 monitor locations and major
roadways with respect to census block group population density
estimates for 2009.
4-38
-------
I : ! I ! ! I
0 40 80 160 Kilometers
Monitor Mean 8-h Daily
Max Ozone, 2007-2009
22 - 44 ppb
45 - 49 ppb
50 - 59 ppb
60 - 69 ppb
70 - 79 ppb
Interstate Highways
Major Highways
Los Angeles CSA Block Groups
2009 Population per Sq Km
0 - 2228
2229 - 4845
^•J 4856 - 8462
^•J 8463 - 14283
^B 15284-27466
• 27467 - 158500
Figure 4-6 Map of the Los Angeles CSA including O$ monitor locations and
major roadways with respect to census block group population
density estimates for 2009.
4-39
-------
Similarities and differences are apparent among the cities. The spatial distribution of
monitor locations in Atlanta and Boston is similar, with one site (site A) near the
high population density area and other monitors in surrounding areas of lower
population density. In Atlanta, the monitors near the city all have similar
concentrations, while somewhat lower concentrations are observed at sites I and J,
which are located >50 km from the city center. Boston shows a different spatial
concentration pattern, with some low and some high concentrations in urban and
less-populated areas. The differences in spatial concentration profiles between the
two cities may be due to more consistent terrain in Atlanta compared with Boston,
which has a coastline, along with the downwind influence of New York and other
northeastern cities contributing to concentration variability.
Los Angeles has a much more complex spatial pattern of monitors, population, and
geography. There are a large number of monitors located in multiple levels of
population density across the entire CSA, which includes substantial rural areas.
Most monitors are near at least moderate population density areas, but there are some
high-density areas without a monitor. Concentrations increase in a somewhat radial
or west-east pattern from the city, with lower concentrations near the port of Long
Beach (monitors B, C, and F). The highest concentrations are located near the San
Bernadino forest (e.g., monitors AG, AO, and AR), which have lower population
density, but more potential for ecological impacts. Low concentrations in highly
populated areas near the coast likely reflect titration by NOX and other atmospheric
constituents, while high downwind concentrations reflect lack of local NOX sources
and increased photochemical processing time.
The location of these monitors relative to the location of dense population centers
varies among urban areas. NCore sites, a subset of the overall monitoring network,
are designed with population exposure as a monitoring objective, and the monitoring
requirements in 40 CFR Part 58, Appendix D include population density as one of
several factors that would be considered in designing the O3 monitoring program for
an area. At least one site for each MSA is designed to be a maximum concentration
site, which could presumably represent the location with the maximum exposure
potential in the city. Sites may also be required upwind and downwind of high-
concentration urban areas.
All three cities have some high population density areas without an O3 monitor.
The siting considerations for NCore monitors generally target the neighborhood
(0.5-4 km) or urban (4-50 km) scale to provide representative concentrations
throughout the metropolitan area; however, a middle-scale (0.1-0.5 km) site may be
acceptable in cases where the site can represent many such locations throughout a
metropolitan area. In other words, a monitor could potentially represent exposures in
other similar areas of the city if land use and atmospheric chemistry conditions are
similar. This is supported by the correlation analyses in Chapter 3.. For example, in
Los Angeles, monitors H and L are located in medium-density areas and show
moderately high correlation (R = 0.78), although they are some 50 km apart.
Although proximity to a monitor does not determine the degree to which that monitor
represents an individual's ambient exposure, it is one indicator. One way to calculate
4-40
-------
monitor representativeness is to calculate the fraction of the urban population living
within a certain radius of a monitor. Table 4-6 presents the fraction of the population
in selected cities living within 1, 5, 10, and 20 km of an O3 monitor. Values are
presented for both total population and for those under 18 years of age, a potentially
susceptible population to the effects of O3. The data indicate that relatively few
people live within 1 km of an O3 monitor, while nearly all of the population in most
cities lives within 20 km of a monitor. Looking at the results for a 5-km radius,
corresponding roughly to the neighborhood scale (Section 3.5.6.1). generally 20-30%
of the population lives within this distance from an O3 monitor. Some cities have a
greater population in this buffer, such as Salt Lake City, while others have a lower
percentage, such as Minneapolis and Seattle. Percentages for children are generally
similar to the total population, with no clear trend.
Another approach is to divide the metropolitan area into sectors surrounding each
monitor such that every person in the sector lives closer to that monitor than any
other. This facilitates calculation of the fraction of the city's population represented
(according to proximity) by each monitor. In Atlanta, for example, the population
fraction represented by each of the 11 monitors in the city ranged from 2.9-22%.
The two monitors closest to the city center (sites A and B on Figure 4-4) accounted
for 16% and 8% of the population, respectively. Site B has two listed monitoring
objectives, highest concentration and population exposure. The other monitor in
Atlanta with a listed objective of highest concentration is Site C, which represents the
largest fraction of the population (22%). The eight monitors with a primary
monitoring objective of population exposure account for 2.9-17% of the population
per monitor.
4-41
-------
Table 4-6 Fraction of the 2009 population living within a specified distance of
an O3 monitor in selected U.S. cities.
Population
City
Atlanta, GA, CSA
Baltimore, MD,
CSA
Birmingham, AL,
CSA
Boston, MA, CSA
Chicago, IL, CSA
Dallas, TX, CSA
Denver, CO, CSA
Detroit, Ml, CSA
Houston, TX, CSA
Los Angeles, CA,
CSA
Minneapolis, MN,
CSA
New York, NY,
CSA
Philadelphia, PA,
CSA
Phoenix, AZ,
CBSA
Pittsburgh, PA,
CSA
Salt Lake City, UT,
CSA
San Antonio, TX,
CBSA
San Francisco, CA,
CSA
Seattle, WA, CSA
St. Louis, MO, CSA
5,
8,
1,
7,
9,
6,
3,
5,
5,
Total
901,670
421,016
204,399
540,533
980,113
791 ,942
103,801
445,448
993,633
18,419,720
3,
652,490
22,223,406
6,
442,836
4,393,462
2,
1,
2,
7,
4,
2,
471 ,403
717,045
061,147
497,443
181,278
914,754
<18yr
1,210,932
1,916,106
281 ,983
1,748,918
2,502,454
1 ,530,877
675,380
1,411,875
1,387,851
4,668,441
872,497
5,284,875
1 ,568,878
873,084
563,309
460,747
484,473
1,675,711
918,309
720,746
Within 1 km
Total
0.3%
1.3%
1.4%
0.9%
1.5%
0.4%
1.7%
0.8%
1.5%
1.6%
0.3%
1.5%
0.9%
2.0%
1.5%
3.0%
0.5%
2.6%
0.3%
1.3%
<18yr
0.3%
1.1%
1.6%
0.9%
1.5%
0.4%
1.6%
0.9%
1.8%
1.7%
0.3%
1.7%
1.0%
2.4%
1.4%
3.0%
0.5%
2.9%
0.3%
1.5%
Within 5 km
Total
8%
25%
22%
17%
28%
13%
35%
15%
26%
28%
5%
23%
22%
35%
22%
41%
12%
41%
5%
17%
<18 yr
9%
24%
24%
16%
29%
13%
36%
17%
28%
29%
4%
23%
24%
41%
21%
38%
12%
40%
5%
18%
Within 10 km
Total
28%
57%
56%
49%
63%
45%
66%
42%
54%
77%
16%
51%
55%
74%
52%
79%
42%
81%
18%
52%
<18 yr
29%
55%
59%
47%
65%
44%
68%
44%
57%
79%
16%
50%
56%
79%
50%
79%
43%
81%
16%
53%
Within 20 km
Total
75%
89%
73%
85%
89%
87%
92%
77%
83%
98%
57%
91%
89%
96%
88%
95%
78%
98%
43%
80%
<18yr
77%
89%
74%
85%
91%
87%
93%
78%
84%
98%
56%
91%
89%
97%
88%
95%
80%
98%
39%
82%
Atlanta population fractions for children (<18 years of age) are similar to those for
the general population, but other populations show a different pattern of monitor
representativeness. Older adults (age 65 and up) were somewhat differently
distributed with respect to the monitors, with most monitors showing a difference of
more than a percentage point compared to the general population. Based on 2000
population data, the fraction of older adults closest to the two city center monitors (A
and B) was 4% higher and 2% lower, respectively, than the fraction for the
population as a whole. Site C showed the highest differential, with 21% of the total
population but only 15% of the older adult population. This indicates the potential for
monitors to differentially represent potentially susceptible populations.
4-42
-------
4.5 Exposure Modeling
In the absence of personal exposure measurements, modeling techniques are used to
estimate exposures, particularly for large populations for which individual-level
measurements would be impractical. Model estimates may be used as inputs to
epidemiologic studies or as stand-alone assessments of the level of exposure likely to
be experienced by a population under certain air quality conditions. This section
describes approaches used to improve exposure estimates, including concentration
surface modeling, which calculates local outdoor concentrations over a geographic
area; air exchange rate modeling, which estimates building ventilation based on
housing characteristics and meteorological parameters; and microenvironment-based
exposure modeling, which combines air quality data with demographic information
and activity pattern simulations to estimate time-weighted exposures based on
concentrations in multiple microenvironments. These models each have strengths and
limitations, as summarized in Table 4-7. The remainder of this section provides more
detail on specific modeling approaches, as well as results of applying the models.
4-43
-------
Table 4-7
Model Type
Concentration
Surface
Air Exchange Rate
Integrated
Microenvironmental
Exposure and Dose
Characteristics of exposure modeling approaches.
Model
Spatial
Interpolation
(e.g., Inverse
Distance
Weighting, Kriging)
Chemistry-
transport
(e.g., CMAQ)
Land-use
regression (LUR)
Mechanistic
(LBL, LBLX)
Empirical
Population
(APEX, SHEDS)
Description
Measured concentrations
are interpolated across an
area to yield local outdoor
concentration estimates
Grid-based O3
concentrations are
calculated from precursor
emissions, meteorology, and
atmospheric chemistry and
physics
Merges concentration data
with local-scale variables
such as land use factors to
yield local concentration
surface
Uses database on building
leakage tests to predict AER
based on building
characteristics and
meteorological variables
(including natural ventilation
in LBLX)
Predicts AER based on
factors such as building age
and floor area
Stochastic treatment of air
quality data, demographic
variables, and activity
pattern to generate
estimates of
microenvironmental
concentrations, exposures,
and doses
Strengths
High concentration
resolution; uses available
data; requires low to
moderate resources for
implementation
First-principles
characterization of
physical and chemical
processes influencing O3
formation
High concentration
resolution
Physical characterization
of driving forces for air
exchange
Low input data
requirements
Probabilistic estimates of
exposure and dose
distributions for specific
populations; consideration
of nonambient sources;
small to moderate
uncertainty for exercising
asthmatic children (APEX)
Limitations
Spatial heterogeneity not
fully captured; a single
high-concentration monitor
can skew results; no
location-activity
information
Grid cell resolution;
resource-intensive; no
location-activity
information
Reactivity and small-scale
spatial variability of O3;
location-specific, limiting
generalizability; no
location-activity
information
Moderate resource
requirement; no location-
activity information
Cannot account for
meteorology; no location-
activity information
Resource-intensive;
evaluation with measured
exposures;
underestimation of
multiple high-exposure
events in an individual
(APEX)
4.5.1 Concentration Surface Modeling
One approach to improve exposure estimates in urban areas involves construction of
a concentration surface over a geographic area, with the concentration at locations
between monitors estimated using a model to compensate for missing data.
The calculated O3 concentration surface can then be used to estimate exposures
outside residences, schools, workplaces, roadways, or other locations of interest. This
technique does not estimate exposure directly because it does not account for activity
patterns or concentrations in different microenvironments. This is an important
consideration in the utility of these methods for exposure assessment; while
improved local-scale estimates of outdoor concentrations may contribute to better
assignment of exposures, information on activity patterns is needed to produce
4-44
-------
estimates of personal exposure. There are three main types of approaches: spatial
interpolation of measured concentrations; statistical models using meteorological
variables, pollutant concentrations, and other predictors to estimate concentrations at
receptors in the domain; and rigorous first-principle models, such as chemistry-
transport models or dispersion models incorporating O3 chemistry. Some researchers
have developed models that combine these techniques. The models may be applied
over urban, regional, or national spatial scales, and can be used to estimate daily
concentrations or longer-term averages. This discussion will focus on short-term
concentrations estimated across urban areas.
The 2006 O3 AQCD (U.S. EPA. 2006b) discussed concentration surface models,
focusing on chemistry-transport models as well as geospatial and spatiotemporal
interpolation techniques (e.g., Christakos and Vyas, 1998a, b; Georgopoulos et al.,
1997). Recent research has continued to refine and extend the modeling approaches.
A few recent papers have compared different approaches for the same urban area.
Marshall et al. (2008) compared four spatial interpolation techniques for estimation
of O3 concentrations in Vancouver, BC. The investigators assigned a daily average
O3 concentration to each of the 51,560 postal-code centroids using one of the
following techniques: (1) the concentration from the nearest monitor within 10 km;
(2) the average of all monitors within 10 km; (3) the inverse-distance-weighted
(IDW) average of all monitors in the area; and (4) the IDW average of the 3 closest
monitors within 50 km. Method 1 (the nearest-monitor approach) and Method 4
(IDW-50 km) had similar mean and median estimated annual- and monthly-average
concentrations, although the 10th-90th percentile range was smaller for IDW-50.
This is consistent with the averaging of extreme values inherent in IDW methods.
The Pearson correlation coefficient between the two methods was 0.93 for monthly-
average concentrations and 0.78 for annual-average concentrations. Methods 2 and 3
were considered sub-optimal and were excluded from further analysis. In the case of
Method 2, a single downtown high-concentration monitor skewed the results in the
vicinity, partially as a result of the asymmetric layout of the coastal city of
Vancouver. Method 3 was too spatially homogenous because it assigned most
locations a concentration near the regional average, except for locations immediately
adjacent to a monitoring site. CMAQ concentration estimates using a 4 km><4 km
grid were also compared to the interpolation techniques in this study. Mean and
median concentrations from CMAQ were approximately 50% higher than Method 1
and Method 4 estimates for both annual and monthly average concentrations. This
may be due in part to the CMAQ grid size, which was too coarse to reveal near-
roadway decrements in O3 concentration due to titration by NO. The IQR for the
annual average was similar between CMAQ and the interpolation techniques, but the
monthly average CMAQ IQR was approximately twice as large, indicating a
seasonal effect.
Bell (2006) compared CMAQ estimates for northern Georgia with nearest-monitor
and spatial interpolation techniques, including IDW and kriging. The area-weighted
concentration estimates from CMAQ indicated areas of spatial heterogeneity that
were not captured by approaches based on the monitoring network. The author
4-45
-------
concluded that some techniques, such as spatial interpolation, were not suitable for
estimation of exposure in certain situations, such as for rural areas. Using the
concentration from the nearest monitor resulted in an overestimation of exposure
relative to model estimates.
Land use regression (LUR) models have been developed to estimate levels of air
pollutants, predominantly NO2, as a function of several land use factors, such as land
use designation, traffic counts, home heating usage, point source strength, and
population density (Ryan and LeMasters, 2007; Gilliland et al., 2005; Briggs et al.,
1997). LUR, initially termed regression mapping (Briggs et al., 1997), is a regression
derived from monitored concentrations as a function of data from a combination of
the land use factors. The regression is then used for predicting concentrations at
multiple locations based on the independent variables at those particular locations
without monitors. Hoek et al. (2008) warn of several limitations of LUR, including
distinguishing real associations between pollutants and covariates from those of
correlated copollutants, limitations in spatial resolution from monitor data,
applicability of the LUR model under changing temporal conditions, and
introduction of confounding factors when LUR is used in epidemiologic studies.
These limitations may partially explain the lack of LUR models that have been
developed for O3 at the urban scale. Brauer et al. (2008) evaluated the use of LUR
and IDW-based spatial-interpolation models in epidemiologic analyses for several
different pollutants in Vancouver, BC and suggested that LUR is appropriate for
directly-emitted pollutants with high spatial variability, such as NO and BC, while
IDW is appropriate for secondary pollutants such as NO2 and PM2.s with less spatial
variability. Although O3 is also a secondary pollutant, its reactivity and high small-
scale spatial variability near high-traffic roadways indicates this conclusion may not
apply for O3.
At a much larger spatial scale, EU-wide, Beelen et al. (2009) compared a LUR model
for O3 with ordinary kriging and universal kriging, which incorporated
meteorological, topographical, and land use variables to characterize the underlying
trend. The LUR model performed reasonably well at rural locations (5-km
resolution), explaining a higher percentage of the variability (R2 = 0.62) than for
other pollutants. However, at the urban scale (1-km resolution), only one variable
was selected into the O3 LUR model (high-density residential land use), and the R2
value was very low (0.06). Universal kriging was the best method for the large-scale
composite EU concentration map, for O3 as well as for NO2 and PMi0, with an R2
value for O3 of 0.70. The authors noted that these methods were not designed to
capture spatial variation in concentrations that are known to occur within tens of
meters of roadways (Section 3.6.2.1), which could partially explain poor model
performance at the urban scale.
Titration of O3 with NO emitted by motor vehicles tends to reduce O3 concentrations
near roadways. McConnell et al. (2006) developed a regression model to predict
residential O3 concentrations in southern California using estimates of residential
NOX calculated from traffic data with the CALINE4 line source dispersion model.
The annual average model results were well-correlated (R2 = 0.97) with multi-year
4-46
-------
average monitoring data. The authors estimated that local traffic contributes 18% of
NOX concentrations measured in the study communities, with the remainder coming
from regional background. Their regression model indicates that residential NOX
reduces residential O3 concentrations by 0.51 ppb (SE 0.11 ppb) O3 per 1 ppb NOX,
and that a 10th-90th percentile increase in local NOX results in a 7.5 ppb decrease in
local O3 concentrations. This intra-urban traffic-related variability in O3
concentrations suggests that traffic patterns are an important factor in the relationship
between central site monitor and residential O3, and that differences in traffic density
between the central site monitor and individual homes could result in either an
overestimate or underestimate of residential O3.
A substantial number of researchers have used geostatistical methods and chemistry-
transport models to estimate O3 concentrations at urban, regional, national, and
continental scales, both in the U.S. and in other countries (Section 3.3). In addition to
short-term exposure assessment for epidemiologic studies, such models may also be
used for long-term exposure assessment, O3 forecasts, or evaluating emission control
strategies. However, as discussed at the beginning of this section, caveats regarding
the importance of activity pattern information in estimating personal and population
exposure should be kept in mind.
4.5.2 Residential Air Exchange Rate Modeling
The residential air exchange rate (AER), which is the airflow into and out of a home,
is an important mechanism for entry of ambient O3. As described in Section 4.3.2.
the indoor-outdoor relationship is greatly affected by the AER. Since studies show
that people spend approximately 66% of their time indoors at home (Leech et al..
2002; Klepeis et al., 2001), the residential AER is a critical parameter for exposure
models, such as APEX, SHEDS, and EMI (discussed in Section 4.5.3) (U.S. EPA.
2011c. 2009b; Burke et al.. 2001). Since the appropriate AER measurements may not
be available for exposure models, mechanistic and empirical (i.e., regression-based)
AER models can be used for exposure assessments. The input data for the AER
models can include building characteristics (e.g., age, number of stories, wind
sheltering), occupant behavior (e.g., window opening), climatic region, and
meteorology (e.g., local temperature and wind speed). Mechanistic AER models use
these meteorological parameters to account for the physical driving forces of the
airflows due to pressure differences across the building envelope from wind and
indoor-outdoor temperature differences (ASHRAE. 2009). Empirical AER models do
not consider the driving forces from the wind and indoor-outdoor temperature
differences. Instead, a scaling constant can be used based on factors such as building
age and floor area (Chan et al.. 2005b).
Single-zone mechanistic models represent a whole-building as a single, well-mixed
compartment. These AER models, such as the Lawrence Berkeley Laboratory (LBL)
model, can predict residential AER using input data from whole-building
pressurization tests (Sherman and Grimsrud. 1980). or leakage area models (Breen et
4-47
-------
al.. 2010; Sherman and Me Williams, 2007). Recently, the LBL air infiltration model
was linked with a leakage area model using population-level census and residential
survey data (Sherman and Me Williams, 2007) and individual-level questionnaire
data (Breen et al., 2010). The LBL model, which predicts the AER from air
infiltration (i.e., small uncontrollable openings in the building envelope) was also
extended to include airflow from natural ventilation (LBLX), and evaluated using
window opening data (Breen et al.. 2010). The AER predictions from the LBL and
LBLX models were compared to daily AER measurements on seven consecutive
days during each season from detached homes in central North Carolina (Breen et al..
2010). For the individual model-predicted and measured AER, the median absolute
difference was 43% (0.17 h'1) and 40% (0.17 h'1) for the LBL and LBLX models,
respectively. Given the uncertainty of the AER measurements (accuracy of 20-25%
for occupied homes), these results demonstrate the feasibility of using these AER
models for both air infiltration (e.g., uncontrollable openings) and natural ventilation
(e.g., window opening) to help reduce the AER uncertainty in exposure models.
The capability of AER models could help support the exposure modeling needs, as
described in Section 4.5.3. which includes the ability to predict indoor concentrations
of ambient O3 that may be substantial for conditions of high AER such as open
windows.
4.5.3 Microenvironment-Based Models
Population-based methods, such as the Air Pollution Exposure (APEX) and
Stochastic Human Exposure and Dose Simulation (SHEDS) integrated
microenvironmental exposure and dose models, involve stochastic treatment of the
model inputs (U.S. EPA. 2009b: Burke et al.. 2001). These are described in detail in
the 2008 NOX ISA (U.S. EPA. 2008c). in AX3.6.1. Stochastic models utilize
distributions of pollutant-related and individual-level variables, such as ambient and
local O3 concentration contributions and breathing rate respectively, to compute the
distribution of individual exposures across the modeled population. The models also
have the capability to estimate received dose through a dosimetry model. Using
distributions of input parameters in the model framework rather than point estimates
allows the models to incorporate uncertainty and variability explicitly into exposure
estimates (Zidek et al.. 2007). These models estimate time-weighted exposure for
modeled individuals by summing exposure in each microenvironment visited during
the exposure period.
The initial set of input data for population exposure models is ambient air quality
data, which may come from a monitoring network or model estimates. Estimates of
concentrations in a set of microenvironments are generated either by mass balance
methods, which can incorporate AER models (Section 4.5.2). or microenvironmental
factors. Microenvironments modeled include indoor residences; other indoor
locations, such as schools, offices, and public buildings; and vehicles. The sequence
of microenvironments and exertion levels during the exposure period is determined
from characteristics of each modeled individual. The APEX model does this by
4-48
-------
generating a profile for each simulated individual by sampling from distributions of
demographic variables such as age, sex, and employment; physiological variables
such as height and weight; and situational variables such as living in a house with a
gas stove or air conditioning. Activity and location (microenvironmental) patterns
from a database such as CHAD are assigned to the simulated individual in a
longitudinal manner, using age, sex, and biometric characteristics (U.S. EPA. 2009a:
Glen et al.. 2008). Breathing rates for each individual are calculated for each activity
based on predicted energy expenditures, and the corresponding dose may then be
computed. APEX has an algorithm to estimate O3 dose and changes in FEVi
resulting from O3 exposure. Summaries of individual- and population-level metrics
are produced, such as maximum exposure or dose, number of individuals exceeding a
specified exposure/dose, and number of person-days at or above benchmark exposure
levels. The models also consider the nonambient contribution to total exposure.
Nonambient source terms are added to the infiltration of ambient pollutants to
calculate the total concentration in the microenvironment. Output from model runs
with and without nonambient sources can be compared to estimate the ambient
contribution to total exposure and dose.
Georgopoulos et al. (2005) used a version of the SHEDS model as the exposure
component of a modeling framework known as MENTOR (Modeling Environment
for Total Risk Studies) in a simulation of O3 exposure in Philadelphia over a 2-week
period in July 1999. Five hundred (500) individuals were sampled from CHAD in
each of 482 census tracts to match local demographic characteristics from U.S.
Census data. Outdoor concentrations over the modeling domain were calculated from
interpolation of photochemical modeling results and fixed-site monitor
concentrations. These concentrations were then used as input data for SHEDS.
Median microenvironmental concentrations predicted by SHEDS for nine simulated
microenvironments were strongly correlated with outdoor concentrations, a result
consistent with the lack of indoor O3 sources in the model. A regression of median
microenvironmental concentrations against outdoor concentrations indicated that the
microenvironmental concentrations were appreciably lower than outdoor
concentrations (regression slope = 0.26). 95th percentile microenvironmental
concentrations were also well correlated with outdoor concentrations and showed a
regression slope of 1.02, although some microenvironmental concentrations were
well below the outdoor values. This suggests that in most cases the high-end
concentrations were associated with outdoor microenvironments. Although the
authors did not report exposure statistics for the population, their dose calculations
also indicated that O3 dose due to time spent outdoors dominated the upper
percentiles of the population dose distribution. They found that both the 50th and
95th percentile O3 concentrations were correlated with census-tract level outdoor
concentrations estimated by photochemical modeling combined with spatiotemporal
interpolation, and attributed this correlation to the lack of indoor sources of O3.
Relationships between exposure and concentrations at fixed-site monitors were not
reported.
An analysis has been conducted for the APEX model to evaluate the contribution of
uncertainty in input parameters and databases to the uncertainty in model outputs
4-49
-------
(Langstaff, 2007). The Monte Carlo analysis indicates that the uncertainty in model
exposure estimates for asthmatic children during moderate exercise is small to
moderate, with 95% confidence intervals of at most ± 6 percentage points at
exposures above 60, 70, and 80 ppb (8-h avg). However, APEX appears to
substantially underestimate the frequency of multiple high-exposure events for a
single individual. The two main sources of uncertainty identified were related to the
activity pattern database and the spatial interpolation of fixed-site monitor
concentrations to other locations. Additional areas identified in the uncertainty
analysis for potential improvement include: further information on children's
activities, including longitudinal patterns in the activity pattern database; improved
information on spatial variation of O3 concentrations, including in near-roadway and
indoor microenvironments; and data from personal exposure monitors with shorter
averaging times to capture peak exposures and lower detection limits to capture low
indoor concentrations. A similar modeling approach has been developed for panel
epidemiologic studies or for controlled human exposure studies, in which activity
pattern data specific to the individuals in the study can be collected. Time-activity
data is combined with questionnaire data on housing characteristics, presence of
indoor or personal sources, and other information to develop a personalized set of
model input parameters for each individual. This model, the Exposure Model for
Individuals, has been developed by EPA's National Exposure Research Laboratory
(U.S. EPA. 20lie: Zartarian and Schultz. 2010).
4.6 Implications for Epidemiologic Studies
Exposure measurement error, which refers to the uncertainty associated with using
exposure metrics to represent the actual exposure of an individual or population, can
be an important contributor to variability in epidemiologic study results. Time-series
studies assess the daily health status of a population of thousands or millions of
people over the course of multiple years (i.e., thousands of days) across an urban area
by estimating their daily exposure using a short monitoring interval (hours to days).
In these studies, the community-averaged concentration of an air pollutant measured
at central-site monitors is typically used as a surrogate for individual or population
ambient exposure. In addition, panel studies, which consist of a relatively small
sample (typically tens) of study participants followed over a period of days to
months, have been used to examine the health effects associated with short-term
exposure to ambient concentrations of air pollutants (Delfino et al., 1996). Panel
studies may also apply a microenvironmental model to represent exposure to an air
pollutant. A longitudinal cohort epidemiologic study, such as the ACS cohort study,
typically involves hundreds or thousands of subjects followed over several years or
decades (Jerrett et al., 2009). Concentrations are generally aggregated over time and
by community to estimate exposures.
Exposure error can under- or over-estimate epidemiologic associations between
ambient pollutant concentrations and health outcomes by biasing effect estimates
toward or away from the null, and tends to widen confidence intervals around those
4-50
-------
estimates (Sheppard et al., 2005; Zeger et al., 2000). Exposure misclassification can
also tend to obscure the presence of potential thresholds for health effects, as
demonstrated by a simulation study of nondifferential exposure misclassification
(Brauer et al., 2002). The importance of exposure misclassification varies with study
design and is dependent on the spatial and temporal aspects of the design. For
example, the use of a community-averaged O3 concentration in a time-series
epidemiologic study may be adequate to represent the day-to-day temporal
concentration variability used to evaluate health effects, but may not capture
differences in the magnitude of exposure due to spatial variability. Other factors that
could influence exposure estimates include nonambient exposure, topography of the
natural and built environment, meteorology, measurement errors, use of ambient O3
concentration as a surrogate for ambient O3 exposure, and the presence of O3 in a
mixture of pollutants. The following sections will consider various sources of error
and how they affect the interpretation of results from epidemiologic studies of
different designs.
4.6.1 Non-Ambient Ozone Exposure
For other criteria pollutants, nonambient sources can be an important contributor to
total personal exposure. There are relatively few indoor sources of O3; as a result,
personal O3 exposure is expected to be dominated by ambient O3 in outdoor
microenvironments and in indoor microenvironments with high air exchange rates
(e.g., with open windows). Even in microenvironments where nonambient exposure
is substantial, such as in a room with an O3 generator, this nonambient exposure is
unlikely to be temporally correlated with ambient O3 exposure (Wilson and Suh,
1997). and therefore would not affect epidemiologic associations between O3 and a
health effect (Sheppard et al.. 2005). In simulations of a nonreactive pollutant,
Sheppard et al. (2005) concluded that nonambient exposure does not influence the
health outcome effect estimate if ambient and nonambient concentrations are
independent. Since personal exposure to ambient O3 is some fraction of the ambient
concentration, it should be noted that effect estimates calculated based on personal
exposure rather than ambient concentration will be increased in proportion to the
ratio of ambient concentration to ambient exposure, and daily fluctuations in this
ratio can widen the confidence intervals in the ambient concentration effect estimate,
but uncorrelated nonambient exposure will not bias the effect estimate (Sheppard et
al.. 2005: Wilson and Suh. 1997).
4.6.2 Spatial and Temporal Variability
Spatial and temporal variability in O3 concentrations can contribute to exposure error
in epidemiologic studies, whether they rely on central-site monitor data or
concentration modeling for exposure assessment. Spatial variability in the magnitude
of concentrations may affect cross-sectional and large-scale cohort studies by
4-51
-------
undermining the assumption that intra-urban concentration and exposure differences
are less important than inter-urban differences. This issue may be less important for
time-series studies, which rely on day-to-day temporal variability in concentrations
to evaluate health effects. Low inter-monitor correlations contribute to exposure error
in time-series studies, including bias toward the null and increased confidence
intervals.
4.6.2.1 Spatial Variability
Spatial variability of O3 concentrations is highly dependent on spatial scale; in effect,
O3 is a regional pollutant subject to varying degrees of local variability. In the
immediate vicinity of roadways, O3 concentrations are reduced due to reaction with
NO and other species (Section 4.3.4.2); over spatial scales of a few kilometers, O3
may be more homogeneous due to its formation as a secondary pollutant; over scales
of tens of kilometers, atmospheric processing can result in higher concentrations
downwind of an urban area than in the urban core. Local-scale variations have a large
impact on the relative magnitude of concentrations among urban monitors, while
conditions favoring high or low rates of O3 formation (e.g., temperature) vary over
large spatial scales. This suggests that neighborhood monitors are likely to track one
another temporally, but miss small-scale spatial variability in magnitude. This is
supported by an analysis in Atlanta, GA, that found correlations greater than 0.8 for
daily O3 concentration metrics (1-h max, 8-h max, and 24-h avg) measured at
monitors 10-60 km apart (Darrow et al., 201 la). In rural areas, a lower degree of
fluctuation in O3 precursors such as NO and VOCs is likely to make the diel
concentration profile less variable than in urban areas, resulting in more sustained
ambient levels. Spatial variability contributes to exposure error if the ambient O3
concentration measured at the central site monitor is used as an ambient exposure
surrogate and differs from the actual ambient O3 concentration outside a subject's
residence and/or worksite (in the absence of indoor O3 sources). Averaging data from
a large number of samplers will dampen intersampler variability, and use of multiple
monitors over smaller land areas may allow for more variability to be incorporated
into an epidemiologic analysis.
Community exposure may not be well represented when monitors cover large areas
with several subcommunities having different sources and topographies, such as the
Los Angeles, CA, CSA (Section 3.6.2.1 and Section 4.4.3). Ozone monitors in
Los Angeles had a much wider range of intermonitor correlations (-0.06 to 0.97) than
Atlanta, GA, (0.61 to 0.96) or Boston, MA, (0.56 to 0.97) using 2007-2009 data.
Although the negative and near-zero correlations in Los Angeles were observed for
monitors located some distance apart (>150 km), some closer monitor pairs had low
positive correlations, likely due to changes in land use, topography, and airflow
patterns over short distances. Lower COD values, which indicate less variability
among monitors in the magnitude of O3 concentrations, were observed in Atlanta
(0.05-0.13) and Boston (0.05-0.19) than Los Angeles (0.05-0.56), although a single
monitor (AM) was responsible for all Los Angeles COD values above 0.40.
4-52
-------
The spatial and temporal variability in O3 concentration in 24 MSAs across the U.S.
was also examined in the 2006 O3 AQCD (U.S. EPA, 2006b) by using Pearson
correlation coefficients, values of the 90th percentile of the absolute difference in O3
concentrations, and CODs. No clear discernible regional differences across the U.S.
were found in the ranges of parameters analyzed.
An analysis of the impact of exposure error due to spatial variability and instrument
imprecision on time-series epidemiologic study results indicated that O3 has
relatively low exposure error compared to other routinely monitored pollutants, and
that the simulated impact on effect estimates is minor. Goldman et al. (2011)
computed population-weighted scaled semivariances and Pearson correlation
coefficients for daily concentration metrics of twelve pollutants measured at multiple
central-site monitors in Atlanta, GA. The 8-h daily max O3 exhibited the lowest
semivariance and highest correlation of any of the pollutants. Although this indicates
some degree of urban-scale homogeneity for O3, the analysis did not account for
near-road effects on O3 concentrations.
Studies evaluating the influence of monitor selection on epidemiologic study results
have found that O3 effect estimates are similar across different spatial averaging
scales and monitoring sites. A study in Italy compared approaches for using fixed-
site monitoring data in a case-crossover epidemiologic study of daily O3 and
mortality (Zauli Sajani et al.. 2011). Ozone effect estimates were found to be similar
whether the nearest monitor was used, or whether single-city, three-city, or six-city
regional averages were used for exposure assessment. In contrast, effect estimates for
PMio and NO2 increased with increasing scale of spatial averaging. Confidence
intervals increased with increasing spatial scale for all pollutants. The authors
attributed the consistency of O3 effect estimates to the relative spatial homogeneity
of O3 over multi-km spatial scales, and pointed to the high (0.85-0.95) inter-monitor
correlations to support this. The use of background monitors rather than monitors
influenced by local sources in this study suggests that local-scale spatial variation in
O3, such as that due to titration by traffic emissions, was not captured in the analyses.
A multi-city U.S. study of asthmatic children found comparable respiratory effect
estimates when restricting the analysis to the monitors closest to the child's zip code
centroid as when using the average of all monitors in the urban area (Mortimer et al.,
2002), suggesting little impact of monitor selection. Sarnat et al. (2010) studied the
spatial variability of O3, along with PM2.5, NO2, and CO, in the Atlanta, GA,
metropolitan area and evaluated how spatial variability affects interpretation of
epidemiologic results, using time-series data for circulatory disease ED visits.
The authors found that associations with ambient 8-h daily max O3 concentration
were similar among all sites tested, including multiple urban sites and a rural site
some 38 miles from the city center. This result was also observed for 24-hour PM2 5
concentrations. In contrast, hourly CO and NO2 showed different associations for the
rural site than the urban sites, although the urban site associations were similar to one
another for CO. This suggests that the choice of monitor may have little impact on
the results of O3 time-series studies, consistent with the moderate to high inter-
monitor correlations observed in Atlanta (Chapter 3).
4-53
-------
One potential explanation for this finding from the study by Sarnat et al. (2010) is
that although spatial variability at different scales contributes to a complicated
pattern of variations in the magnitude of O3 concentrations between near-road, urban
core, and urban downwind sites, day-to-day fluctuations in concentrations may be
reflected across multiple urban microenvironments. In addition, time-averaging of O3
and PM2.5 concentrations may smooth out some of the intra-day spatial variability
observed with the hourly CO and NO2 concentrations. However, some uncertainty in
observed effect estimates due to spatial variability and associated exposure error is
expected to remain, including a potential bias toward the null.
4.6.2.2 Seasonality
The relationship between personal exposure and ambient concentration has been
found to vary by season, with at least three factors potentially contributing to this
variation: differences in building ventilation (e.g., air conditioning or heater use
versus open window ventilation), higher O3 concentrations during the O3 season
contributing to increased exposure and improved detection by personal monitors; and
changes in activity pattern resulting in more time spent outside. Evidence has been
presented in studies conducted in several cities regarding the effect of ventilation on
personal-ambient and indoor-outdoor O3 relationships (see Section 4.3.2 and
Section 4.3.3). More limited evidence is available regarding the specific effects of O3
detection limits and activity pattern changes on O3 relationships.
Several studies have found increased summertime correlations or ratios between
personal exposure and ambient concentration (Sarnat et al.. 2005: Sarnat et al.. 2000)
or between indoor and outdoor O3 concentrations (Geyh et al.. 2000: Avol et al..
1998a). However, others have found higher ratios in fall than in summer (Sarnat et
al., 2006a) or equivalent, near-zero ratios in winter and summer (Sarnat et al., 2001),
possibly because summertime use of air conditioners decreases building air exchange
rates. It should be noted that O3 concentrations during winter are generally much
lower than summertime concentrations, possibly obscuring wintertime relationships
due to detection limit issues. Studies specifically evaluating the effect of ventilation
conditions on O3 relationships have found increased correlations or ratios for
individuals or buildings experiencing higher air exchange rates (Sarnat et al., 2006a:
Gevh et al.. 2000: Sarnat et al.. 2000: Romieuetal.. 1998a).
Increased correlations or ratios between personal exposure and ambient
concentration, or between indoor and outdoor concentration, are likely to reduce
error in exposure estimates used in epidemiologic studies. This suggests that studies
conducted during the O3 season or in periods when communities are likely to have
high air exchange rates (e.g., during mild weather) may be less prone to exposure
error than studies conducted only during winter. Year-round studies that include both
the O3 and non-O3 seasons may have an intermediate level of exposure error.
4-54
-------
4.6.3 Exposure Duration
Epidemiologic studies of health effects associated with short-term and long-term
exposures use different air pollution metrics and thus have different sources of
exposure error. The following subsections discuss the impact of using different short-
term and long-term exposure metrics on epidemiologic results.
4.6.3.1 Short-Term Exposure
The averaging time of the daily exposure metrics used to evaluate daily aggregated
health data (e.g., 1-h or 8-h daily max, versus 24-h avg concentration) may also
impact epidemiologic results, since different studies report different daily metrics.
Correlations between 1-h daily max, 8-h daily max, and 24-h avg concentrations for
U.S. monitoring sites are presented in Section 3.6.1 (Figure 3-23 and accompanying
text). The two daily peak values (1-h max and 8-h max) are well correlated, with a
median (IQR) correlation of 0.97 (0.96-0.98). The correlation between the 8-h max
and 24-h avg are somewhat less well correlated with a median (IQR) correlation of
0.89 (0.86-0.92). While this may complicate quantitative comparisons between
epidemiologic studies using different daily metrics, as well as the interpretation of
studies using metrics other than the current 8-hour standard, the high inter-metric
correlations suggest it is a relatively small source of uncertainty in comparing the
results of studies using different metrics. This is supported by a study comparing
each of these metrics in a time-series study of respiratory ED visits (Darrow et al.
2011 a), which found positive associations for all metrics, with the strongest
association for the 8-h daily max exposure metric (Section 6.2.7.3).
The ratios of 1-h daily max, 8-h daily max, and 24-h avg concentrations to one
another have been found to differ across communities and across time within
individual communities (Anderson and Bell. 2010). For example, 8:24 hour ratios
ranged from 1.23-1.83, with a median of 1.53. Lower ratios were generally observed
in the spring and summer compared to fall and winter. Ozone concentration was
identified as the most important predictor of O3 metric ratios, with higher overall O3
concentrations associated with lower ratios. In communities with higher long-term
O3 concentrations, the lower 8:24 hour ratio is attributed to high baseline O3, which
results in elevated 24-h average values. Differences in the representativeness of O3
metrics introduces uncertainty into the interpretation of epidemiologic results and
complicates comparison of studies using different metrics. Preferably, studies will
report results using multiple metrics. In cases where this does not occur, the results of
the study by Anderson and Bell (2010) can inform the uncertainty associated with
using a standard increment to adjust effect estimates based on different metrics so
that they are comparable (Chapter 6).
A study compared measures of spatial and temporal variability for 1-h daily max and
24-h daily avg O3 concentrations in Brazil (Bravo and Bell, 2011). The 1-h daily
max value was found to have higher correlation between monitors (i.e., lower
4-55
-------
temporal variability) and lower COD (a measure of spatiotemporal variability which
incorporates differences in concentration magnitude, with lower values indicating
lower variability; see Chapter 1) than the 24-h avg value. The range of correlation
coefficients and COD values was similar between the two metrics, although the
variation was lower for the 1-h daily max, as indicated by the R2 value for the
regression of correlation coefficient on inter-monitor distance.
4.6.3.2 Long-Term Exposure
A study in Canada suggests that an exposure metric based on a single year can
represent exposure over a multi-decade period. The authors compared exposure
assessment methods for long-term O3 exposure and found that the annual average
concentration in the census tract of a subject's residence during 1980 and 1994 was
well-correlated (0.76 and 0.82, respectively) with a concentration metric accounting
for movement among census subdivisions during 1980-2002 (Guay et al., 2011). This
may have been due in part to a relatively low rate of movement, with subjects
residing on average for 71% of the 22-year period in the same census subdivision
they were in during 1980.
Analysis of the exposure assessment methodology in a recent study of mortality
associated with long-term O3 exposure (Jerrett et al.. 2009) is illustrative. In this
study, the authors computed quarterly averages of the daily 1-h max O3
concentration, averaged the two summer quarters together to produce an annual
value, then calculated a 23-year average value for each city in the study. Producing a
single value for each city enables a comparison of relatively cleaner cities with
relatively more polluted cities. In this case, the average was calculated using the 1-h
daily max value; if the 24-h avg value had been used, concentrations would have
been lower and potentially more variable, based on analyses in Chapter 3_. According
to Table 3-7, the 2007-2009, 3-year average 1-h daily max value during the warm
season was approximately 50% higher than the corresponding 24-avg value on a
nationwide basis. Correlation between the two metrics varies by site, indicating the
differential influence of the overnight period on 24-h avg concentrations. The median
correlation between 1-h daily max and 24-h avg is 0.83, with an IQR of 0.78-0.88.
It is not clear, however, that a different exposure assignment method would yield
different results.
Long-term O3 trends, as discussed in Chapter 3_, show gradually decreasing
concentrations. Figure 3-48 shows that concentrations have decreased most for the
90th percentile, with relatively little change among the 10th percentile monitors.
The decrease has been greater in the eastern U.S. than in the western part of the
country (excluding California). For the most part, the rank order of regions in terms
of O3 concentration has remained the same, as shown in Figure 3-50. with the
Northeast, Southeast, and California exhibiting the highest concentrations.
The decreasing trend is consistent across nearly all monitors in the U.S., with only
1-2% of monitors reporting an increase of more than 5 ppb between the 2001-2003
4-56
-------
and 2008-2010 time periods (Figure 3-52 and Figure 3-53). This figure provides
some evidence that epidemiologic studies of long-term exposure are not affected by
drastic changes in O3 concentration, such as a relatively clean city becoming highly
polluted or the reverse.
A few epidemiologic studies have evaluated the impact of distance to monitor on
associations between long-term O3 concentration and reproductive outcomes, as
discussed in Chapter 7. It is not clear from this evidence whether using a local
monitor for these multi-month concentration averages improves exposure
assessment. Jalaludin et al. (2007) found somewhat higher effect estimates for
women living within 5 km of a fixed-site O3 monitor than for all women in the
Sydney, Australia, metropolitan area, suggesting that increased monitor proximity
reduced exposure misclassification. In contrast, Darrow et al. (201 Ib) found no
substantial difference between effect estimates for those living within 4 mi of a
fixed-site monitor and those living in the five-county area around Atlanta, GA. This
result could be due to spatial variability over smaller scales than the 4-mi radius
evaluated, time spent away from the residence impacting O3 exposure, or similarity
in monitor location and representativeness across the urban area (see Figure 4-4).
At this time, the effect of exposure error on long-term exposure epidemiologic
studies is unclear.
4.6.4 Relationship between Personal Exposure and Ambient
Concentration
Personal exposure is generally moderately correlated with ambient O3 concentration,
although the magnitude of personal exposures is often much lower than the
magnitude of ambient concentrations (Section 4.3.3). Moderate correlation between
personal exposure and ambient concentration indicates that concentration-based
exposure metrics are capturing the variability in exposure needed for epidemiologic
studies, particularly for time-series and panel studies. Low personal-ambient
correlations reported in the literature are strongly influenced by high detection limits
of personal samplers. This results in a high fraction of personal samples below the
detection limit that include substantial random variation and are thus unable to
provide a signal that could correlate with variations in ambient concentration. Low
correlations in situations with a high proportion of samples below the detection limit
should not be interpreted as evidence for the lack of a relationship between personal
exposure and ambient O3 concentrations. To the extent that true correlations are less
than one, epidemiologic effect estimates based on ambient concentration will be
biased toward the null (Zeger et al.. 2000). High detection limits are less of an issue
for ratios of personal exposure to ambient concentration, for which a low personal
sample value likely represents an actual low exposure, and thus appropriately leads
to a low ratio. Low ratios result from low penetration and high reaction of O3 in
indoor microenvironments where people spend most of their time. This results in
attenuation of the magnitude of the exposure-based effect estimate or response
function relative to the ambient concentration-based response function
4-57
-------
(see Equation 4-5), although if the ratio is approximately constant over time, the
strength of the statistical association would be similar for concentration- and
exposure-based effect stimates (Sheppard, 2005; Sheppard et al., 2005).
In addition to the effect of the correlations and ratios themselves, spatial variation in
their values across urban areas also impacts epidemiologic results. In this case, the
exposure error is not likely to cause substantial bias, but tends more toward widening
confidence intervals, thus reducing the precision of the effect estimate (Zeger et al.,
2000). This loss of precision is due to the Berkson-like nature of this spatial
variation, in which individual or subpopulation correlations and ratios tend to vary
about the overall population mean.
Long-term O3 exposure studies are not available that permit evaluation of the
relationship between long-term O3 concentrations and personal or population
exposure. The value of short-term exposure data for evaluating long-term
concentration-exposure relationships is uncertain. If the longer averaging time
(annual, versus daily or hourly) smooths out short-term fluctuations, long-term
concentrations may be well-correlated with long-term exposures. However, lower
correlation between long-term exposures and ambient concentration could occur if
important exposure determinants change over a period of several years, including
activity pattern and residential air exchange rate.
4.6.5 Exposure to Copollutants and Ozone Reaction Products
Although indoor O3 concentrations are usually well below ambient concentrations,
the same reactions that reduce O3 indoors form particulate and gaseous species,
including other oxidants, as summarized in Section 4.3.4.3. Exposures to these
reaction products would therefore be expected to be correlated with ambient O3
concentrations, although no evidence was identified regarding personal exposures.
Such exposure could potentially contribute to health effects observed in
epidemiologic studies.
4.6.6 Averting Behavior
As described in Section 4.4.2, several recent studies indicate that some lifestages and
populations alter their behavior on high O3 days to avoid exposure. Such behavioral
responses to information about forecasted air quality may introduce systematic
measurement error in air pollution exposure, leading to biased estimates of the
impact of air pollution on health. For example, studies have hypothesized that
variation in time spent outdoors may be a driving factor behind the considerable
heterogeneity in O3 mortality impacts across communities (Bell et al., 2004).
If averting behavior reduces outdoor O3 exposure, then studies that do not account
for averting behavior may produce effect estimates that are biased toward the null
(Section 6.2.7.2).
4-58
-------
This is supported by an epidemiologic study that examined the association between
exposure to ambient O3 concentrations and asthma hospitalizations in Southern
California during 1989-1997, which indicates that controlling for avoidance behavior
increases the effect estimate for both children and older adults, but not for adults
aged 20-64 (Neidell andKinnev. 2010: Neidell 2009). Figure 4-7 and Figure 4-8.
reproduced from Neidell (2009). show covariate-adjusted asthma hospital admissions
as a function of daily maximum 1-hour O3 concentration for all days (gray line) and
days when no O3 alert was issued (black line). Stage 1 smog alerts were issued by the
State of California for days when ambient O3 concentrations were forecast to be
above 0.20 ppm; however, the concentration-response functions are based on
measured O3 concentrations. For children aged 5-19 (Figure 4-7). hospital
admissions were higher on high-O3 days when no alert was issued, especially on
days with O3 concentrations above 0.15 ppm (150 ppb). The concentration-response
curves for all days and days with no alert diverge at measured O3 concentrations
between 0.10 and 0.15 ppm because smog alerts begin to be issued more frequently
in this range. This suggests that in the absence of information that would enable
averting behavior, children experience higher O3 exposure and subsequently a
greater number of asthma hospital admissions than on alert days with similar O3
concentrations. The lower rate of admissions observed when alert days were included
in the analysis suggests that averting behavior reduced O3 exposure and asthma
hospital admissions. In both cases, O3 was found to be associated with asthma
hospital admissions, although the strength of the association is underestimated when
not accounting for averting behavior. A different result was observed when
examining associations for adults aged 20-64 (Figure 4-8). who had similar rates of
hospital admissions on non-alert days as on all days. The lack of change for adults
aged 20-64, which is primary employment age, may reflect lower response to air
quality alerts due to the increased opportunity cost of behavior change. The finding
that air quality information reduces the daily asthma hospitalization rate in these
populations provides additional support for a link between O3 and health effects.
4-59
-------
A
.05
,15 .2
O/one (ppm)
Overall . No Alert
.25
Source: Reprinted with permission of the Board of Regents of the University of Wisconsin System, University of Wisconsin Press
(Neidell. 2009).
Figure 4-7 Adjusted asthma hospital admissions by age on lagged O3 by alert
status, ages 5-19 years old.
B
.05
O/one (ppm)
.NoAli-rt
Source: Reprinted with permission of the Board of Regents of the University of Wisconsin System, University of Wisconsin Press
(Neidell. 2009).
Figure 4-8 Adjusted asthma hospital admissions by age on lagged O3 by alert
status, ages 20-64 years old.
4-60
-------
4.6.7 Exposure Estimation Methods in Epidemiologic Studies
Epidemiologic studies use a variety of methods to assign exposure. Study design,
data availability, and research objectives are all important factors for epidemiologists
when selecting an exposure assessment method. Common methods for assigning
exposure using monitoring data include using a single fixed-site monitor to represent
population exposure, averaging concentrations from multiple monitors, and selecting
the closest monitor. Investigators may also use statistical adjustment methods, such
as trimming extreme values, to prepare the concentration data set. Panel or small-
scale cohort studies involving dozens of individuals may use more individualized
concentration measurements, such as personal exposures, residential indoor or
outdoor measurements, or concentration data from local study-specific monitors. For
long-term epidemiologic studies, the lack of personal exposure data or dedicated
measurements means that investigators must rely on fixed-site monitoring data.
These data may be used directly, averaged across counties or other geographic areas,
or used to construct geospatial or regression models to assign concentrations to
unmonitored locations. Longer-term averages (months to years) are typically used
(e.g., in studies discussed in Section 7.3.1.1). Chapters 6 and 7 describe the exposure
assessment methods used in the epidemiologic studies described therein, providing
additional detail on studies with innovative or expanded techniques designed to
improve exposure assessment and reduce exposure error.
The use of O3 measurements from central ambient monitoring sites is the most
common method for assigning exposure in epidemiologic studies. However, fixed-
site measurements do not account for the effects of spatial variation in O3
concentration, ambient and non-ambient concentration differences, and varying
activity patterns on personal exposures (Brown et al., 2009; Chang et al., 2000; Zeger
et al., 2000). Inter-individual variability in exposure error across a population will be
minimal when: (1) O3 concentrations are uniform across the region; (2) personal
activity patterns are similar across the population; and (3) housing characteristics,
such as air exchange rate and indoor reaction rate, are constant over the study area.
To the extent that these factors vary by location and population, there will be errors
in the magnitude of total exposure based solely on ambient monitoring data.
Modeled concentrations can also be used as exposure surrogates in epidemiologic
studies, as discussed in Section 4.5. Geostatistical spatial interpolation techniques,
such as IDW and kriging, can provide finer-scale estimates of local concentration
over urban areas. A microenvironmental modeling approach simulates exposure
using empirical distributions of concentrations in specific microenvironments
together with human activity pattern data. The main advantage of the modeling
approach is that it can be used to estimate exposures over a wide range of population
and scenarios. However, this probabilistic, distribution-based approach is not well-
suited to estimate exposures for specific individuals, such as might be needed for
4-61
-------
cohort or panel epidemiologic studies. Another main disadvantage of the modeling
approach is that the results of modeling exposure assessment must be compared to an
independent set of measured exposure levels (Klepeis, 1999). In addition,
resource-intensive development of validated and representative model inputs is
required, such as human activity patterns, distributions of air exchange rate, and
deposition rate. Therefore, modeled exposures are used much less frequently in
epidemiologic studies.
4.7 Summary and Conclusions
This section will briefly summarize and synthesize the main points of the chapter,
with particular attention to the relevance of the material for the interpretation of
epidemiologic studies.
Passive badge samplers are the most widely used technique for measuring personal
O3 exposure (Section 4.3.1). The detection limit of the badges for a 24-hour
sampling period is approximately 5-10 ppb, with lower detection limits at longer
sampling durations. In low-concentration conditions this may result in an appreciable
fraction of 24-hour samples being below the detection limit. The use of more
sensitive portable active monitors, including some that have recently become
available, may help overcome this issue and improve personal monitoring in the
future.
Since there are relatively few indoor sources of O3, indoor O3 concentrations are
often substantially lower than outdoor concentrations due to reactions of O3 with
indoor surfaces and airborne constituents (Section 4.3.2). Air exchange rate is a key
determinant of the I/O ratio, which is generally in the range of 0.1-0.4 (Table 4-1).
with some evidence for higher ratios during the O3 season when concentrations are
higher.
Personal exposure is moderately correlated with ambient O3 concentration, as
indicated by studies reporting correlations generally in the range of 0.3-0.8
(Table 4-2). Hourly concentration correlations are more variable than those averaged
over 24 hours or longer, with correlations in outdoor microenvironments (0.7-0.9)
much higher than those in residential indoor (0.1) or other indoor (0.3-0.4)
microenvironments. Some studies report substantially lower personal-ambient
correlations, a result attributable in part to low air exchange rate and O3
concentrations below the sampler detection limit, conditions often encountered
during wintertime. Low correlations may also occur for individuals or populations
spending substantial time indoors.
The ratio between personal exposure and ambient concentration varies widely
depending on activity patterns, housing characteristics, and season, with higher
personal-ambient ratios generally observed with increasing time spent outside, higher
air exchange rate, and in seasons other than winter (Table 4-3). Personal-ambient
ratios are typically 0.1-0.3, although individuals spending substantial time outdoors
4-62
-------
(e.g., outdoor workers) may have much higher ratios (0.5-0.9). Low personal-
ambient ratios result in attenuation of the magnitude of the exposure-based effect
estimate or response function relative to the concentration-based response function,
although the statistical association is similar for concentration- and exposure-based
effect estimates if the ratio is approximately constant over time.
Personal exposure to other pollutants shows variable association with personal
exposure to O3, with differences in copollutant relationships depending on factors
such as season, city-specific characteristics, activity pattern, and spatial variability of
the copollutant (Section 4.3.4). In near-road and on-road microenvironments,
correlations between O3 and traffic-related pollutants are moderately to strongly
negative, with the most strongly negative correlations observed for NO2 (-0.8 to
-0.9). This is consistent with the chemistry of NO oxidation, in which O3 is
consumed to form NO2. The more moderate negative correlations observed for
PM2.s, PMi.o, and VOC may reflect reduced concentrations of O3 in polluted
environments due to other scavenging reactions. A similar process occurs indoors,
where infiltrated O3 reacts with airborne or surface-associated materials to form
secondary compounds, such as formaldehyde. Although such reactions decrease
indoor O3 exposure, they result in increasing exposure to other species which may
themselves have health effects.
Variations in ambient O3 concentrations occur over multiple spatial and temporal
scales. Near roadways, O3 concentrations are reduced due to reaction with NO and
other species (Section 4.3.4.2). Over spatial scales of a few kilometers and away
from roads, O3 may be somewhat more homogeneous due to its formation as a
secondary pollutant, while over scales of tens of kilometers, additional atmospheric
processing can result in higher concentrations downwind of an urban area. Although
local-scale variability impacts the magnitude of O3 concentrations, O3 formation
rates are influenced by factors that vary over larger spatial scales, such as
temperature (Section 3.2), suggesting that urban monitors may track one another
temporally but miss small-scale variability in magnitude. The resulting uncertainty in
exposure contributes to exposure measurement error in epidemiologic studies.
Another factor that may influence epidemiologic results is the tendency for people to
avoid O3 exposure by altering their behavior (e.g., reducing time spent outdoors) on
high-O3 days. Activity pattern has a substantial effect on ambient O3 exposure, with
time spent outdoors contributing to increased exposure (Section 4.4.2). Averting
behavior has been predominantly observed among children, older adults, and people
with respiratory problems. Such effects are less pronounced in the general
population, possibly due to the opportunity cost of behavior modification. Evidence
from one recent epidemiologic study indicates increased asthma hospital admissions
among children and older adults when O3 alert days were excluded from the analysis
(presumably thereby eliminating averting behavior based on high O3 forecasts).
The lower rate of admissions observed when alert days were included in the analysis
suggests that estimates of health effects based on concentration-response functions
which do not account for averting behavior may be biased toward the null.
4-63
-------
The range of personal-ambient correlations reported by most studies (0.3-0.8) is
similar to that for NO2 (U.S. EPA, 2008c) and somewhat lower than that for PM2.5
(U.S. EPA, 2009d). To the extent that relative changes in central-site monitor
concentration are associated with relative changes in exposure concentration, this
indicates that ambient monitor concentrations are representative of day-to-day
changes in average total personal exposure and in personal exposure to ambient O3.
The lack of indoor sources of O3, in contrast to NO2 and PM2.5, is partly responsible
for low indoor-outdoor ratios (generally 0.1-0.4) and low personal-ambient ratios
(generally 0.1-0.3), although it contributes to increased personal-ambient
correlations. The lack of indoor sources also suggests that fluctuations in ambient O3
may be primarily responsible for changes in personal exposure, even under low-
ventilation, low-concentration conditions. Nevertheless, low personal-ambient
correlations are a source of exposure error for epidemiologic studies, tending to
obscure the presence of potential thresholds, bias effect estimates toward the null,
and widen confidence intervals, and this impact may be more pronounced among
populations spending substantial time indoors. The impact of this exposure error may
tend more toward widening confidence intervals than biasing effect estimates, since
epidemiologic studies evaluating the influence of monitor selection indicate that
effect estimates are similar across different spatial averaging scales and monitoring
sites.
4-64
-------
References
Akinbami. LJ; Lynch. CD; Parker. JD; Woodruif. TJ. (2010). The association between childhood asthma
prevalence and monitored air pollutants in metropolitan areas, United States, 2001-2004. Environ Res 110:
294-301. http://dx.doi.0rg/10.1016/i.envres.2010.01.001
Anderson. GB; Bell. ML. (2010). Does one size fit all? The suitability of standard ozone exposure metric
conversion ratios and implications for epidemiology. J Expo Sci Environ Epidemiol 20: 2-11.
http://dx.doi.org/10.1038/ies.2008.69
Anderson. SE; Wells. JR: Fedorowicz. A; Butterworth. LF; Meade. BJ; Munson. AE. (2007). Evaluation of the
contact and respiratory sensitization potential of volatile organic compounds generated by simulated indoor
air chemistry. Toxicol Sci 97: 355-363. http://dx.doi.org/10.1093/toxsci/kfm043
Aoki. T; Tanabe. S. (2007). Generation of sub-micron particles and secondary pollutants from building
materials by ozone reaction. Atmos Environ 41: 3139-3150.
http://dx.doi.0rg/10.1016/i.atmosenv.2006.07.053
ASHRAE (American Society of Heating, Refrigerating and Air-Conditioning Engineers). (2009). The 2009
ASHRAE handbook: Fundamentals. Atlanta, GA: American Society of Heating, Refrigerating and Air-
Conditioning Engineers Inc. http://www.ashrae.org/publications/page/2283
Avol. EL; Navidi. WC: Colome. SD. (1998a). Modeling ozone levels in and around southern California homes.
Environ Sci Technol 32: 463-468.
Beckerman. B; Jerrett. M; Brook. JR; Verma. DK; Arain. MA; Finkelstein. MM. (2008). Correlation of
nitrogen dioxide with other traffic pollutants near a major expressway. Atmos Environ 42: 275-290.
http://dx.doi.0rg/10.1016/i.atmosenv.2007.09.042
Beelen. R; Hoek. G: Pebesma. E; Vienneau. D; de Hoogh. K; Briggs. DJ. (2009). Mapping of background air
pollution at a fine spatial scale across the European Union. Sci Total Environ 407: 1852-1867.
http://dx.doi.0rg/10.1016/i.scitotenv.2008.ll.048
Beko. G: Clausen. G: Weschler. CJ. (2007). Further studies of oxidation processes on filter surfaces: Evidence
for oxidation products and the influence of time in service. Atmos Environ 41: 5202-5212.
http://dx.doi.0rg/10.1016/i.atmosenv.2006.07.063
Bell. ML. (2006). The use of ambient air quality modeling to estimate individual and population exposure for
human health research: A case study of ozone in the Northern Georgia region of the United States. Environ
Int32: 586-593.
Bell. ML; McDermott. A; Zeger. SL; Samet. JM; Dominici. F. (2004). Ozone and short-term mortality in 95
US urban communities, 1987-2000. JAMA 292: 2372-2378. http://dx.doi.org/10.1001/iama.292.19.2372
Blanken. PD; Dillon. J; Wismann. G. (2001). The impact of an air quality advisory program on voluntary
mobile source air pollution reduction. Atmos Environ 35: 2417-2421.
Blondeau. P; lordache. V: Poupard. O: Genin. D; Allard. F. (2005). Relationship between outdoor and indoor
air quality in eight French schools. Indoor Air 15: 2-12.
Brauer. M; Brook. JR. (1997). Ozone personal exposures and health effects for selected groups residing in the
Fraser Valley. Atmos Environ 31: 2113-2121.
Brauer. M; Brumm. J: Vedal. S: Petkau. AJ. (2002). Exposure misclassification and threshold concentrations in
time series analyses of air pollution health effects. Risk Anal 22: 1183-1193.
Brauer. M; Lencar. C: Tamburic. L; Koehoorn. M; Demers. P; Karr. C. (2008). A cohort study of traffic-related
air pollution impacts on birth outcomes. Environ Health Perspect 116: 680-686.
http://dx.doi.org/10.1289/ehp.10952
4-65
-------
Bravo. MA; Bell. ML. (2011). Spatial heterogeneity of PM10 and O3 in Sao Paulo, Brazil, and implications
for human health studies. J Air Waste Manag Assoc 61: 69-77.
Breen. MS: Breen. M; Williams. RW: Schultz. BD. (2010). Predicting residential air exchange rates from
questionnaires and meteorology: Model evaluation in central North Carolina. Environ Sci Technol 44:
9349-9356. http://dx.doi.org/10.1021/esl01800k
Bresnahan. BW: Dickie. M; Gerking. S. (1997). Averting behavior and urban air pollution. Land Econ 73: 34-
57.
Briggs. DJ: Collins. S: Elliott. P; Fischer. P; Kingham. S: Lebret. E; Pryl. K; Van Reeuwijk. H; Smallbone. K;
Van Per Veen. A. (1997). Mapping urban air pollution using GIS: A regression-based approach. Int J Geogr
Inform Sci 11: 699-718.
Brown. K; Sarnat. J: Suh. H; Coull. B; Koutrakis. P. (2009). Factors influencing relationships between personal
and ambient concentrations of gaseous and paniculate pollutants. Sci Total Environ 407: 3754-3765.
http://dx.doi.0rg/10.1016/i.scitotenv.2009.02.016
Burke. JM; Zufall. MJ; Ozkaynak. H. (2001). A population exposure model for paniculate matter: Case study
results for PM2.5 in Philadelphia, PA. J Expo Sci Environ Epidemiol 11: 470-489.
Chan. WR: Nazaroff. WW: Price. PN; Sohn. MD; Gadgil. AJ. (2005b). Analyzing a database of residential air
leakage in the United States. Atmos Environ 39: 3445-3455.
Chang. LT; Koutrakis. P: Catalano. PJ; Suh. HH. (2000). Hourly personal exposures to fine particles and
gaseous pollutants-Results from Baltimore, Maryland. J Air Waste Manag Assoc 50: 1223-1235.
Chen. X; Hopke. PK; Carter. WP. (2011). Secondary organic aerosol from ozonolysis of biogenic volatile
organic compounds: Chamber studies of particle and reactive oxygen species formation. Environ Sci
Technol 45: 276-282. http://dx.doi.org/10.1021/esl02166c
Christakos. G: Vvas. VM. (1998a). A composite space/time approach to studying ozone distribution over
eastern United States. Atmos Environ 32: 2845-2857. http://dx.doi.org/10.1016/S1352-2310(98)00407-5
Christakos. G: Vvas. VM. (1998b). A novel method for studying population health impacts of spatiotemporal
ozone distribution. Soc Sci Med 47: 1051-1066.
Darrow. LA; Klein. M; Sarnat. JA; Mulholland. JA; Strickland. MJ: Sarnat. SE; Russell. AG: Tolbert. PE.
(2011 a). The use of alternative pollutant metrics in time-series studies of ambient air pollution and
respiratory emergency department visits. J Expo Sci Environ Epidemiol 21: 10-19.
http://dx.doi.org/10.1038/ies.2009.49
Darrow. LA; Klein. M; Strickland. MJ: Mulholland. JA: Tolbert. PE. (20 lib). Ambient air pollution and birth
weight in full-term infants in Atlanta, 1994-2004. Environ Health Perspect 119: 731-737.
http://dx.doi.org/10.1289/ehp.1002785
Delfino. PJ: Coate. BD; Zeiger. RS: Seltzer. JM; Street. DH; Koutrakis. P. (1996). Daily asthma severity in
relation to personal ozone exposure and outdoor fungal spores. Am J Respir Crit Care Med 154: 633-641.
Gauderman. WJ: Avol. E; Gilliland. F; Vora. H; Thomas. D; Berhane. K; McConnell. R; Kuenzli. N; Lurmann.
F; Rappaport. E; Margolis. H; Bates. D; Peters. J. (2004). The effect of air pollution on lung development
from 10 to 18 years of age. N Engl J Med 351: 1057-1067.
Georgopoulos. PG: Purushothaman. V; Chiou. R. (1997). Comparative evaluation of methods for estimating
potential human exposure to ozone: Photochemical modeling and ambient monitoring. J Expo Sci Environ
Epidemiol 7: 191-215.
Georgopoulos. PG: Wang. SW: Vvas. VM; Sun. Q; Burke. J: Vedantham. R; McCurdv. T; Ozkaynak. H.
(2005). A source-to-dose assessment of population exposures to fine PM and ozone in Philadelphia, PA,
during a summer 1999 episode. J Expo Sci Environ Epidemiol 15: 439-457.
Gevh. AS: Roberts. PT; Lurmann. FW: Schoell. BM; Avol. EL. (1999). Initial field evaluation of the Harvard
active ozone sampler for personal ozone monitoring. J Expo Sci Environ Epidemiol 9: 143-149.
4-66
-------
Geyh. AS: Wolfson. JM; Koutrakis. P; Mulik. JD; Avol. EL. (1997). Development and evaluation of a small
active ozone sampler. Environ Sci Technol 31: 2326-2330.
Geyh. AS: Xue. J: Ozkaynak. H; Spengler. JD. (2000). The Harvard Southern California chronic ozone
exposure study: Assessing ozone exposure of grade-school-age children in two southern California
communities. Environ Health Perspect 108: 265-270.
Gilliland. F; Avol. E; Kinnev. P; Jerrett M; Dvonch. T; Lurmann. F; Buckley. T; Brevsse. P; Keeler. G: de
Villiers. T; Mcconnell. R. (2005). Air pollution exposure assessment for epidemiologic studies of pregnant
women and children: Lessons learned from the Centers for Children's Environmental Health and Disease
Prevention Research. Environ Health Perspect 113: 1447-1454. http://dx.doi.org/10.1289/ehp.7673
Glen. G: Smith. L; Isaacs. K; Mccurdv. T; Langstaff. J. (2008). Anew method of longitudinal diary assembly
for human exposure modeling. J Expo Sci Environ Epidemiol 18: 299-311.
http://dx.doi.org/10.1038/sj.jes.7500595
Goldman. GT: Mulholland. JA: Russell. AG: Strickland. MJ: Klein. M: Waller. LA: Tolbert. PE. (2011).
Impact of exposure measurement error in air pollution epidemiology: Effect of error type in time-series
studies. Environ Health Global Access Sci Source 10: 61. http://dx.doi.org/10.1186/1476-069X-10-61
Grosjean. D; Hisham. MWM. (1992). A passive sampler for atmospheric ozone. J Air Waste Manag Assoc 42:
169-173.
Guav. M; Stieb. DM; Smith-Doiron. M. (2011). Assessment of long-term exposure to air pollution in a
longitudinal national health survey. J Expo Sci Environ Epidemiol 21: 337-342.
http://dx.doi.org/10.1038/ies.2010.37
Henry. GT; Gordon. CS. (2003). Driving less for better air: Impacts of a public information campaign. J Policy
Anal Manage 22: 45-63. http://dx.doi.org/10.1002/pam.10095
Heroux. ME; Clark. N; Van Ryswyk. K; Mallick. R; Gilbert. NL; Harrison. I; Rispler. K; Wang. D;
Anastassopoulos. A; Guav. M; MacNeill. M; Wheeler. AJ. (2010). Predictors of indoor air concentrations in
smoking and non-smoking residences. Int J Environ Res Public Health 7: 3080-3099.
http://dx.doi.org/10.3390/ijerph7083080
Hoek. G: Beelen. R; de Hoogh. K; Vienneau. D: Gulliver. J: Fischer. P; Briggs. D. (2008). A review of land-
use regression models to assess spatial variation of outdoor air pollution [Review]. Atmos Environ 42:
7561-7578.
Hyttinen. M; Pasanen. P; Kalliokoski. P. (2006). Removal of ozone on clean, dusty and sooty supply air filters.
Atmos Environ 40: 315-325.
Jalaludin. B; Mannes. T; Morgan. G: Lincoln. D; Sheppeard. V: Corbett. S. (2007). Impact of ambient air
pollution on gestational age is modified by season in Sydney, Australia. Environ Health 6: 16.
http://dx.doi.org/10.1186/1476-069X-6-16
Jerrett. M; Burnett. RT; Pope. CA. Ill; Ito. K; Thurston. G: Krewski. D; Shi. Y; Calle. E; Thun. M. (2009).
Long-term ozone exposure and mortality. N Engl J Med 360: 1085-1095.
http://dx.doi.org/10.1056/NEJMoa0803894
Kanno. S: Yanagisawa. Y. (1992). Passive ozone/oxidant sampler with coulometric determination using
iodine/nylon-6 charge-transfer complex. Environ Sci Technol 26: 744-749.
http://dx.doi.org/10.1021/es00028a012
Karner. AA; Eisinger. DS: Niemeier. DA. (2010). Near-roadway air quality: Synthesizing the findings from
real-world data. Environ Sci Technol 44: 5334-5344. http://dx.doi.org/10.1021/esl00008x
Klepeis. NE. (1999). An introduction to the indirect exposure assessment approach: Modeling human exposure
using microenvironmental measurements and the recent National Human Activity Pattern Survey. Environ
Health Perspect 107: 365-374.
Klepeis. NE; Nelson. WC: Ott. WR; Robinson. JP; Tsang. AM; Switzer. P; Behar. JV: Hern. SC: Engelmann.
WH. (2001). The National Human Activity Pattern Survey (NHAPS): A resource for assessing exposure to
environmental pollutants. J Expo Anal Environ Epidemiol 11: 231-252.
4-67
-------
Klepeis. NE; Tsang. AM; Behar. JV. (1996). Analysis of the national human activity pattern survey (NHAPS)
respondents from a standpoint of exposure assessment [EPA Report]. (EPA/600/R-96/074). Washington,
DC: U.S. Environmental Protection Agency.
Koutrakis. P; Suh. HH; Sarnat. JA; Brown. KW: Coull. BA; Schwartz. J. (2005). Characterization of
paniculate and gas exposures of sensitive subpopulations living in Baltimore and Boston (pp. 1-65;
discussion 67-75). (131). Boston, MA: Health Effects Institute.
http://pubs.healtheffects.org/view.php?id=91
Koutrakis. P; Wolfson. JM; Bunyaviroch. A; Froehlich. SE; Hirano. K; Mulik. JD. (1993). Measurement of
ambient ozone using a nitrite-coated filter. Anal Chem 65: 209-214.
Langstaff. JE. (2007). Analysis of uncertainty in ozone population exposure modeling [technical
memorandum]. Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards.
Lee. K; Parkhurst. WJ: Xue. J: Ozkaynak. H; Neuberg. D; Spengler. JD. (2004a). Outdoor/indoor/personal
ozone exposures of children in Nashville, Tennessee. J Air Waste Manag Assoc 54: 352-359.
Leech. JA; Nelson. WC: Burnett. RT; Aaron. S: Raizenne. ME. (2002). It's about time: A comparison of
Canadian and American time-activity patterns. J Expo Anal Environ Epidemiol 12: 427-432.
http://dx.doi.org/10.1038/sj.jea.7500244
Liard. R; Zureik. M; Le Moullec. Y; Soussan. D; Glorian. M; Grimfeld. A; Neukirch. F. (1999). Use of
personal passive samplers for measurement of NO2, NO, and O3 levels in panel studies. Environ Res 81:
339-348.
Linn. WS: Shamoo. DA; Anderson. KR; Peng. RC: Avol. EL; Hackney. JD: Gong. H. Jr. (1996). Short-term air
pollution exposures and responses in Los Angeles area schoolchildren. J Expo Sci Environ Epidemiol 6:
449-472.
Liu. LJS: Koutrakis. P: Leech. J: Broder. I. (1995). Assessment of ozone exposures in the greater metropolitan
Toronto area. J Air Waste Manag Assoc 45: 223-234.
Lopez-Aparicio. S: Smolik. J: Maskova. L; Souckova. M; Grantoft. T; Ondrackova. L; Stankiewicz. J. (2011).
Relationship of indoor and outdoor air pollutants in a naturally ventilated historical building envelope.
Build Environ 46: 1460-1468. http://dx.doi.Org/10.1016/i.buildenv.2011.01.013
Mansfield. CA; Johnson. FR; Van Houtven. GL. (2006). The missing piece: Valuing averting behavior for
children's ozone exposures. Resource Energ Econ 28: 215-228.
http://dx.doi.0rg/10.1016/i.reseneeco.2006.02.002
Marshall. JD: Netherv. E; Brauer. M. (2008). Within-urban variability in ambient air pollution: Comparison of
estimation methods. Atmos Environ 42: 1359-1369. http://dx.doi.Org/10.1016/i.atmosenv.2007.08.012
McConnell. R; Berhane. K; Yao. L; Lurmann. FW: Avol. E; Peters. JM. (2006). Predicting residential ozone
deficits from nearby traffic. Sci Total Environ 363: 166-174.
http://dx.doi.0rg/10.1016/i.scitotenv.2005.06.028
McCurdv. T; Glen. G: Smith. L; Lakkadi. Y. (2000). The National Exposure Research Laboratory's
consolidated human activity database. J Expo Sci Environ Epidemiol 10: 566-578.
McDermott. M; Srivastava. R: Croskell. S. (2006). Awareness of and compliance with air pollution advisories:
A comparison of parents of asthmatics with other parents. J Asthma 43: 235-239.
htrp://dx.doi.org/10.1080/02770900600567114
McDonnell. WF; Abbey. DE; Nishino. N; Lebowitz. MD. (1999a). Long-term ambient ozone concentration
and the incidence of asthma in nonsmoking adults: the Ahsmog study. Environ Res 80: 110-121.
Mortimer. KM; Neas. LM; Dockerv. DW: Redline. S: Tager. IB. (2002). The effect of air pollution on inner-
city children with asthma. Eur Respir J 19: 699-705. http://dx.doi.org/10.1183/09031936.02.00247102
Neidell. M. (2009). Information, avoidance behavior, and health: The effect of ozone on asthma
hospitalizations. Journal of Human Resources 44: 450-478.
4-68
-------
Neidell. M. (2010). Air quality warnings and outdoor activities: Evidence from Southern California using a
regression discontinuity design. J Epidemiol Community Health 64: 921-926.
http://dx.doi.org/10.1136/iech.2008.081489
Neidell. M; Kinnev. PL. (2010). Estimates of the association between ozone and asthma hospitalizations that
account for behavioral responses to air quality information. Environ Sci Pol 13: 97-103.
http://dx.doi.0rg/10.1016/i.envsci.2009.12.006
O'Neill. MS: Ramirez-Aguilar. M; Meneses-Gonzalez. F; Hernandez-Avila. M; Geyh. AS: Sienra-Monge. JJ:
Romieu. I. (2003). Ozone exposure among Mexico City outdoor workers. J Air Waste Manag Assoc 53:
339-346.
Ogawa & Co (Ogawa & Company). (2007). Ambient air passive sampler for NO-NO2, NOx, SO2, O3, NH3.
Pompano Beach, FL: Ogawa & Company USA, Inc. http://www.ogawausa.com/passive.html
Ramirez-Aguilar. M; Barraza-Villarreal. A; Moreno-Macias. H; Winer. AM; Cicero-Fernandez. P; Velez-
Marquez. MG: Cortez-Lugo. M; Sienra-Monge. JJ: Romieu. I. (2008). Assessment of personal exposure to
ozone in asthmatic children residing in Mexico City. Salud Publica Mex 50: 67-75.
Reiss. R; Ryan. PB; Koutrakis. P; Tibbetts. SJ. (1995a). Ozone reactive chemistry on interior latex paint.
Environ Sci Technol 29: 1906-1912.
Reiss. R; Ryan. PB; Tibbetts. SJ: Koutrakis. P. (1995b). Measurement of organic acids, aldehydes, and ketones
in residential environments and their relation to ozone. J Air Waste Manag Assoc 45: 811-822.
Riediker. M; Williams. R: Devlin. R; Griggs. T; Bromberg. P. (2003). Exposure to particulate matter, volatile
organic compounds, and other air pollutants inside patrol cars. Environ Sci Technol 37: 2084-2093.
http://dx.doi.org/10.1021/es026264v
Romieu. I; Lugo. MC: Colome. S: Garcia. AM; Avila. MH; Geyh. A; Velasco. SR; Rendon. EP (1998a).
Evaluation of indoor ozone concentration and predictors of indoor-outdoor ratio in Mexico City. J Air
Waste Manag Assoc 48: 327-335.
Ryan. PH; LeMasters. GK. (2007). A review of land-use regression models for characterizing intraurban air
pollution exposure [Review]. Inhal Toxicol 19: 127.
Sarnat. JA; Brown. KW: Schwartz. J: Coull. BA; Koutrakis. P. (2005). Ambient gas concentrations and
personal particulate matter exposures: Implications for studying the health effects of particles.
Epidemiology 16: 385-395. http://dx.doi.org/10.1097/01.ede.0000155505.04775.33
Sarnat. JA; Koutrakis. P: Suh. HH. (2000). Assessing the relationship between personal particulate and gaseous
exposures of senior citizens living in Baltimore, MD. JAir Waste Manag Assoc 50: 1184-1198.
Sarnat. JA: Schwartz. J: Catalano. PJ; Suh. HH. (2001). Gaseous pollutants in particulate matter epidemiology:
Confounders or surrogates? Environ Health Perspect 109: 1053-1061.
Sarnat. SE; Coull. BA; Schwartz. J: Gold. PR; Suh. HH. (2006a). Factors affecting the association between
ambient concentrations and personal exposures to particles and gases. Environ Health Perspect 114: 649-
654.
Sarnat. SB; Klein. M; Sarnat. JA: Flanders. WD; Waller. LA; Mulholland. JA: Russell. AG: Tolbert. PE.
(2010). An examination of exposure measurement error from air pollutant spatial variability in time-series
studies. J Expo Sci Environ Epidemiol 20: 135-146. http://dx.doi.org/10.1038/ies.2009.10
Semenza. JC: Wilson. DJ; Parra. J: Bontempo. BD; Hart. M; Sailor. DJ; George. LA. (2008). Public perception
and behavior change in relationship to hot weather and air pollution. Environ Res 107: 401-411.
http://dx.doi.0rg/10.1016/i.envres.2008.03.005
Sheppard. L. (2005). Acute air pollution effects: consequences of exposure distribution and measurements. J
Toxicol Environ Health A 68: 1127-1135.
Sheppard. L; Slaughter. JC: Schildcrout. J: L-JS. L; Lumlev. T. (2005). Exposure and measurement
contributions to estimates of acute air pollution effects. J Expo Sci Environ Epidemiol 15: 366-376.
4-69
-------
Sherman. M; McWilliams. J. (2007). Air leakage of U.S. homes: Model prediction. (LBNL-62078). Berkeley,
CA: Lawrence Berkeley National Laboratory, http://epb.lbl.gov/publications/lbnl-62078.pdf
Sherman. MH; Grimsrud. DT. (1980). Infiltration-pressurization correlation: Simplified physical modeling.
In ASHRAE Transactions. Berkeley, CA: Lawrence Berkeley Laboratory. http://eetd.lbl.gov/ie/pdf/LBL-
10163.pdf
Shu. S: Morrison. GC. (2011). Surface reaction rate and probability of ozone and alpha-terpineol on glass,
polyvinyl chloride, and latex paint surfaces. Environ Sci Technol 45: 4285-4292.
http://dx.doi.org/10.1021/es200194e
Suh. HH; Zanobetti. A. (2010). Exposure error masks the relationship between traffic-related air pollution and
heart rate variability [Erratum]. J Occup Environ Med 52: 1138.
http://dx.doi.org/10.1097/JOM.Ob013e3181fd2632
U.S. EPA (U.S. Environmental Protection Agency). (2006c). Estimating contributions of outdoor fine particles
to indoor concentrations and personal exposures: Effects of household characteristics and personal
activities [EPA Report]. (EPA/600/R-06/023). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (2008c). Integrated science assessment for oxides of
nitrogen: Health criteria [EPA Report]. (EPA/600/R-08/071). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=194645
U.S. EPA (U.S. Environmental Protection Agency). (2009a). Consolidated Human Activity Database.
Available online at http://www.epa.gov/chadnetl/ (accessed August 27, 2009).
U.S. EPA (U.S. Environmental Protection Agency). (2009b). Human exposure modeling: Air pollutants
exposure model (APEX/TRIM.Expo Inhalation). Available online at
http://www.epa.gov/ttn/fera/human apex.html (accessed June 13, 2012).
U.S. EPA (U.S. Environmental Protection Agency). (2009d). Integrated science assessment for particulate
matter [EPA Report]. (EPA/600/R-08/139F). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=216546
U.S. EPA (U.S. Environmental Protection Agency). (2011 a). AirNow. Available online at
http://www.airnow.gov/ (accessed January 28, 2011).
U.S. EPA (U.S. Environmental Protection Agency). (2011b). Exposure factors handbook 2011 edition (final)
[EPA Report]. (EPA/600/R-09/052F). http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=236252
U.S. EPA (U.S. Environmental Protection Agency). (2011c). Exposure model for individuals. Available online
at http://www.epa.gov/heasd/products/emi/emi.html (accessed June 11, 2012).
Wen. XJ; Balluz. L; Mokdad. A. (2009). Association between media alerts of air quality index and change of
outdoor activity among adult asthma in six states, BRFSS, 2005. J Community Health 34: 40-46.
http://dx.doi.org/10.1007/sl0900-008-9126-4
Weschler. CJ: Shields. HC. (1997). Potential reactions among indoor pollutants. Atmos Environ 31: 3487-
3495.
White. IR: Martin. D; Mufioz. MP; Petersson. FK; Henshaw. SJ: Nickless. G: Lloyd-Jones. GC: Clemitshaw.
KG: Shallcross. DE. (2010). Use of reactive tracers to determine ambient OH radical concentrations:
Application within the indoor environment. Environ Sci Technol 44: 6269-6274.
http://dx.doi.org/10.1021/es901699a
Williams. R; Rea. A; Vette. A; Croghan. C: Whitaker. D; Stevens. C: Mcdow. S: Fortmann. R; Sheldon. L;
Wilson. H; Thornburg. J: Phillips. M; Lawless. P; Rodes. C: Daughtrev. H. (2009b). The design and field
implementation of the Detroit exposure and aerosol research study. J Expo Sci Environ Epidemiol 19: 643-
659. http://dx.doi.org/10.1038/jes.2008.61
Wilson. KL; Birks. JW. (2006). Mechanism and elimination of a water vapor interference in the measurement
of ozone by UV absorbance. Environ Sci Technol 40: 6361-6367. http://dx.doi.org/10.1021/es052590c
4-70
-------
Wilson. WE; Mage. DT; Grant. LD. (2000). Estimating separately personal exposure to ambient and
nonambient paniculate matter for epidemiology and risk assessment: Why and how. J Air Waste Manag
Assoc 50: 1167-1183.
Wilson. WE; Suh. HH. (1997). Fine particles and coarse particles: Concentration relationships relevant to
epidemiologic studies. JAir Waste Manag Assoc 47: 1238-1249.
Xue. J: Liu. SV: Ozkavnak. H; Spengler. JD. (2005). Parameter evaluation and model validation of ozone
exposure assessment using Harvard Southern California Chronic Ozone Exposure Study data. J Air Waste
Manag Assoc 55: 1508-1515.
Xue. J: McCurdv. T; Spengler. J: Ozkavnak. H. (2004). Understanding variability in time spent in selected
locations for 7-12-year old children. J Expo Anal Environ Epidemiol 14: 222-233.
http://dx.doi.org/10.1038/si.iea.7500319
Zartarian. VG: Schultz. BD. (2010). The EPA's human exposure research program for assessing cumulative
risk in communities. J Expo Sci Environ Epidemiol 20: 351-358. http://dx.doi.org/10.1038/ies.2009.20
Zauli Saiani. S: Hanninen. O; Marchesi. S: Lauriola. P. (2011). Comparison of different exposure settings in a
case-crossover study on air pollution and daily mortality: Counterintuitive results. J Expo Sci Environ
Epidemiol 21: 385-394. http://dx.doi.org/10.1038/jes.2010.27
Zeger. SL; Thomas. D; Dominici. F; Samet. JM; Schwartz. J: Dockery. D; Cohen. A. (2000). Exposure
measurement error in time-series studies of air pollution: Concepts and consequences. Environ Health
Perspect 108: 419-426.
Zidek. JV: Shaddick. G: Meloche. J: Chatfield. C: White. R. (2007). A framework for predicting personal
exposures to environmental hazards. Environ Ecol Stat 14: 411-431.
4-71
-------
5 DOSIMETRY, MODE OF ACTION, AND SPECIES
HOMOLOGY
5.1 Introduction
This chapter has three main purposes. The first is to describe the principles that
underlie the dosimetry of O3 and to discuss factors that influence it. The second is to
describe the modes of action leading to the health effects that will be presented in
Chapters 6 and 7. The third is to discuss the homology of responses in animals and
humans exposed to O3 and the interspecies differences that may affect these
responses. This chapter is not intended to be a comprehensive overview, but rather, it
updates the basic concepts derived from O3 literature presented in previous
documents (U.S. EPA. 2006b. 1996a) and introduces the recent relevant literature.
In Section 5.2, particular attention is given to dosimetric factors influencing
individual risk of developing effects from O3 exposure. As there have been few O3
dosimetry studies published since the last AQCD, the reader is referred to previous
documents (U.S. EPA, 2006b, 1996a) for more detailed discussion of the past
literature. Evaluation of the progress in the interpretation of past dosimetry studies,
as well as studies published since 2005, in the areas of uptake, reactions, and models
for O3 dosimetry, is discussed.
Section 5.3 highlights findings of studies published since the 2006 O3 AQCD, which
provide insight into the biological pathways by which O3 exerts its actions. Since
common mechanisms lead to health effects from both short- and long-term exposure
to O3, these pathways are discussed in Chapter 5 rather than in later chapters.
The related sections of Chapters 6 and 7 are indicated. Earlier studies that represent
the current state of the science are also discussed. Studies conducted at more
environmentally-relevant concentrations of O3 are of greater interest, since
mechanisms responsible for effects at low O3 concentrations may not be identical to
those occurring at high O3 concentrations. Some studies at higher concentrations are
included if they were early demonstrations of key mechanisms or if they are recent
demonstrations of potentially important new mechanisms. The topics of dosimetry
and mode of action are bridged by reactions of O3 with components of the
extracellular lining fluid (ELF), which play a role in both O3 uptake and biological
responses (Figure 5-1).
In addition, this chapter discusses interindividual variability in responses, and issues
related to species comparison of doses and responses (Section 5.4 and Section 5.5).
These topics are included in this chapter because they are influenced by both
dosimetric and mechanistic considerations.
5-1
-------
O3 exposure
Inhaled
O,dose
Net
O3dose
Tissue O3 dose
and product
formation
Modes of Action
Health Effects
Note: Ozone transport follows a path from exposure concentration, to inhaled dose, to net dose, to the local tissue dose.
Chapter 5 discusses the concepts of dose and modes of action that result in the health effects discussed in Chapters 6 and 7.
Figure 5-1 Schematic of the Os exposure and response pathway.
5.2 Human and Animal Ozone Dosimetry
5.2.1 Introduction
Dosimetry refers to the measurement or estimation of the quantity of or rate at which
a chemical and/or its reaction products are absorbed and retained at target sites.
Figure 5-1 illustrates the transport of O3 or its reaction products from exposure to
dose to the development of health effects. Ozone exposure has been defined in
Section 4.2 and consists of contact between the human or animal and O3 at a specific
concentration for a specified period of time (i.e., exposure = concentration x time).
The amount of O3 present in a given volume of air for which animals and individuals
are exposed is termed exposure concentration. Ozone exposure will result in some
amount (dose) of O3 crossing an exposure surface to enter a target area. The initial
measure of dose after O3 enters the respiratory tract (RT) is inhaled dose and is the
amount or rate of O3 that crosses the outer RT surface before crossing the ELF and is
effectively C x t xVE, where C is concentration, t is time, and VE is minute
ventilation. Ozone may then cross from the gas phase across the ELF interface where
net dose may be measured. Net dose is the amount or rate of entry of O3 across the
gas/ELF interface. In modeling studies, the dose rate is often expressed as a flux per
unit of surface area of a region of respiratory epithelium. Finally, O3 or its reaction
products may reach the tissues and tissue dose of O3 can be reported. Tissue dose is
the amount of O3 or its reaction products absorbed and available for reacting with
tissues and is difficult and rarely measured. In the literature, the exposure
concentration and various measures of dose (i.e., net dose and inhaled dose) are often
used as surrogates for tissue dose. However, ambient or exposure concentrations are
not a true measure of dose so understanding the relationship between ambient
5-2
-------
concentrations and tissue dose allows for a greater appreciation of the dose-response
from O3 exposure.
Posterior
Nasal Passage
Nasal Part
Tracheobronchial
Region
Bronchiolar Region
Bronchioles
Terminal Bronchioles
Respiratory Bronchioles
Alveolar Interstitial
Alveolar Duct +
Alveoli
Note: Structures are anterior nasal passages, ET-i; oral airway and posterior nasal passages, ET2; bronchial airways, BB;
bronchioles, bb; and alveolar interstitial, Al.
Source: Based on ICRP (1994).
Figure 5-2 Representation of respiratory tract regions in humans.
Ozone is a highly reactive, though poorly water soluble, gas at physiological
temperature. The latter feature is believed to be the reason why it is able to penetrate
into targets in the lower respiratory tract (LRT). Figure 5-2 presents the basic
structure of the human RT. The lung can be divided into three major regions: the
extrathoracic (ET) region or upper respiratory tract (URT, from the nose/mouth to
the end of the larynx); the tracheobronchial (TB) tree (from trachea to the terminal
bronchioles); and the alveolar or pulmonary region (from the respiratory bronchioles
to the terminal alveolar sacs). The latter two regions comprise the LRT. Although the
structure varies, the illustrated anatomic regions are common to all mammalian
5-3
-------
species with the exception of the respiratory bronchioles. Respiratory bronchioles,
the transition region between ciliated and fully alveolated airways, are found in
humans, dogs, ferrets, cats, and monkeys. Respiratory bronchioles are absent in rats
and mice and abbreviated in hamsters, guinea pigs, sheep, and pigs. The branching
structure of the ciliated bronchi and bronchioles also differs between species from
being a rather symmetric and dichotomous branching network of airways in humans
to a more monopodial branching network in other mammals.
Figure 5-3 illustrates the structure of the LRT with progression from the large
airways in the TB region to the alveolus in the alveolar region. The fact that O3 is so
chemically reactive has suggested to some that its tissue dose at the target sites exists
in the form of oxidation products such as aldehydes and peroxides (see
Section 5.2.3). Reaction products are formed when O3 interacts with components of
the ELF such as lipids and antioxidants. The ELF varies throughout the length of the
RT with the nasal airways through the bronchial tree lined with a thicker layer of
ELF than the alveolar region (Figure 5-3b). Ozone dose is directly related to the
coupled diffusion and chemical reactions occurring in ELF, a process termed
"reactive absorption." Thus, the O3 dose depends on both the concentration of O3 as
well as the availability of substrates within the ELF.
Ozone dose is affected by complex interactions between a number of other major
factors including RT morphology, breathing route, frequency, and volume,
physicochemical properties of the gas, physical processes of gas transport, as well as
the physical and chemical properties of the ELF and tissue layers (Figure 5-3 c).
The role of these processes varies throughout the length of the RT and as O3 moves
from the gas to liquid compartments of the RT.
5-4
-------
b.
c.
Air
Lquic
nun
Tissue
Tissue
Air
Air
Transport Factors
Gas Phase
Convection
Diffusion
Dispersion
Liquid Phase
Solubility
Diffusion
Chemical Reaction
Convection
Note: (a) Illustrates basic airway anatomy. Structures are epithelial cells, EP; basement membrane, BM; smooth muscle cells, SM;
and fibrocartilaginous coat, FC. (b) Illustrates the relative amounts of liquid, tissue, and blood with distal progression. In the
bronchi there is a thick surface lining over a relatively thick layer of tissues. With distal progress, the lining diminishes allowing
increased access of compounds crossing the air-liquid interface to the tissues and the blood, (c) Presents the factors acting in the
gas and liquid phases of O3 transport.
Source: Panel (a) reprinted with permission of McGraw-Hill (Weibel. 1980).
Figure 5-3 Structure of lower airways with progression from the large airways
to the alveolar region.
Two types of measurements have been used to arrive at the O3 dose to target sites
during breathing: (1) measurement of removal of O3 from the air stream (termed
"uptake"); and (2) measurement of chemical reactions in tissues or with
biomolecules known to be present in tissues (termed "reactants"). The results of the
above measurements have been incorporated into mathematical models for the
purpose of explaining, predicting, and extrapolating O3 dose in different exposure
5-5
-------
scenarios. Few new studies have investigated the uptake of O3 in the RT since the
last O3 assessment (U.S. EPA, 2006b). The studies that have been conducted
generally agree with the results presented in the past and do not change the dosimetry
conclusions of the last document.
5.2.2 Ozone Uptake
Past AQCDs provide information on the majority of literature relevant to
understanding the state of the science in O3 dosimetry. Measurements of O3 dose
have been inferred from simultaneous measurements of airflow and O3 concentration
at the airway opening of the nose or mouth (Nodelman and Ultman, 1999; Wiester et
al., 1996a) as well as at internal sampling catheters (Gerrity et al., 1995; Gerrity et
al., 1988). One method of quantifying O3 dose is to measure the amount of O3
removed from the air stream during breathing (termed "uptake"). The difference in
the amount of O3 inhaled and exhaled relative to the amount of inhaled O3 is termed
fractional absorption. Uptake efficiency is also reported and refers to the O3 absorbed
in a region expressed as a fraction of the total amount of O3 entering the given
region. Uptake studies have utilized bolus and continuous O3 breathing techniques as
well as modeling to investigate these measures of uptake and the distribution of O3
uptake between the URT and LRT. A number of the studies that have measured the
fractional absorption and uptake efficiency of O3 in the human RT, URT, and LRT
are presented in Table 5-1. For studies that reported fractional absorption of O3
boluses, the equivalent fractional absorption of a continuous inhalation of O3 was
estimated as the sum of the products of the experimental bolus absorption and
incremental volume of a bolus into a breath divided by the tidal volume of the breath,
or, where available, was taken from Table 1A of Schlesinger et al. (1997).
5-6
-------
Table 5-1
Human respiratory tract uptake efficiency data.
Inspiratory
Mouth/ Flow
Reference Nose3 (mL/sec)
URT,
fs complete
VT (mL) (bpm) breath
Uptake Efficiency
LRT, Total RT,
URT, complete tidal
inspiration breath breath
Continuous Exposure
Gerritv et al.
(1 988)
Gerritv et al.
(1 994)
Gerritv et al.
(1995)
Wiester et al.
(1 996a)
OR
N
OR/N
OR/N
OR/N
OR
OR
OR
OR
N
Face mask
Santiago et al.
(2001)
N
N
509
456
500
350
634
1,360
1,360
330
539
514
480
50
250
832 18
754 18
800 18
832 12
778 24
1 ,650 25
1 ,239 35
825 12
631 16
642 16
1,100 27.6
0.40 0.91
0.36 0.91
0.43 0.91
0.41 0.93
0.38 0.89
0.37 0.43 0.81
0.41 0.36 0.78
0.27 0.95 0.91
0.76
0.73
0.86
0.80°
0.33
Bolus Exposure
Huetal.
(1 992)
Kabel et al.
(1 994)
Huetal.
(1 994)
Ultman et al.
(1 994)
Bush et al.
(1 996)
Mouth-
piece
Mouth-
piece
Mouth-
piece
N
Mouth-
piece
Mouth-
piece
Mouth-
piece
Mouth-
piece
Mouth-
piece
Mouth-
piece
Mouth-
piece
Mouth-
piece
250
250
250
250
150
250
500
750
1,000
250
250
250
500 0.46
500 0.50
500 0.53
500 0.78
500 0.65
500 0.51
500 0.26
500 0.16
500 0.11
500d 15 0.30
500 15 0.47
500 0.51
0.88
0.88
0.88
0.94
0.91
0.87
0.82
0.78
0.76
0.89
5-7
-------
Reference
Nodelman and
Ultman (1999)
Ultman et al.
(2004)
Mouth/
Nose3
Nasal
Cannula
Nasal
Cannula
Mouth-
piece
Mouth-
piece
OR
OR
Inspiratory
Flow
(mL/sec)
150
1,000
150
1,000
490
517
VT (mL)
500
500
500
500
450d
574
fs
(bpm)b
18
120
18
120
32.7
27
URT,
complete
breath
0.90
0.50
0.77
0.25
Uptake Efficiency
LRT, Total RT,
URT, complete tidal
inspiration breath breath
0.92
0.84
0.91
0.75
0.87
0.91
aOR = oral exposure during spontaneous breathing; N = nasal exposure during spontaneous breathing; OR/N = pooled data from
oral and nasal exposure; mouthpiece = exposure by mouthpiece.
bfB is either measured or is computed from flows and VT.
0 FURT from Santiago et al. (2001) represents nasal absorption (Fnose).
dVT is computed from flow and fB.
5.2.2.1 Gas Transport Principles
The three-dimensional transport of O3 in the lumen of an airway is governed by
diffusion associated with the Brownian motion of gas molecules and convection that
depends on local velocity patterns. Simultaneously, O3 is absorbed from the gas
stream into the ELF where it undergoes simultaneous radial diffusion and chemical
reaction.
When air flows through an airway, O3 located near the tube center moves faster than
O3 near the tube wall where frictional forces retard the flow. This non-uniformity in
the radial profile of velocity gives rise to an axial spreading or dispersion of the O3
that operates in parallel with bulk flow and axial diffusion The shape of the velocity
profile is affected by the flow direction through bifurcating airway branches
(Schroter and Sudlow. 1969). The velocity profile is nearly parabolic during
inhalation but quite flat during exhalation. Thus, there tends to be greater axial
dispersion during inhalation than during exhalation. Dispersion also depends on the
nature of the flow, that is, whether it is laminar (i.e., streamlined) or turbulent
(i.e., possessing random velocity fluctuations). Because turbulent flow flattens
velocity profiles, it may actually diminish dispersion. In humans, turbulent flow
persists only a few generations into the RT. The persistence of turbulence into the RT
also varies by species and flow rates. For example, airflow is nonturbulent in the rat
nose at any physiologic flow rate but may be highly turbulent in the human nose
during exercise (Miller. 1995).
5-8
-------
The relative importance of axial convection, diffusion, and dispersion varies among
RT regions for a given level of ventilation. In the URT and major bronchi, axial
convection and dispersion tend to be the predominant mechanisms. Moving into
more distal areas of the RT, the summed cross-sectional area of the airways rapidly
increases and linear velocities decrease, leading to a greater role for molecular
diffusion. The principal mechanism of gas mixing in the lung periphery is molecular
diffusion (Engel. 1985).
Absorption of O3 at the airway wall depends on a concentration boundary layer on
the gas side of the airway wall as well as simultaneous radial diffusion and chemical
reaction within the ELF (Figure 5-3c) (Miller, 1995). The boundary layer caused by
slowly moving gas near the airway wall can be an important component of the radial
diffusion resistance to O3 absorption. This diffusive resistance increases with distal
penetration into the RT with one study reporting that the gas boundary layer
contributes 53% of the overall diffusive resistance in the URT, 78% in the proximal
LRT, and 87% in the distal LRT (Hu et al., 1994). The geometry of airway surfaces
also affects local O3 absorption. For example, nasal and lung regions receive
different O3 exposures or doses (Miller and Kimbell, 1995); and larger surface-to-
volume ratio of the smaller airways in women enhances local O3 uptake and reduces
the distal penetration volume of O3 into the RT of women relative to men (TJltman et
al.. 2004).
5.2.2.2 Target Sites for Ozone Dose
A primary uptake site of O3 delivery to the lung epithelium is believed to be the
centriacinar region (CAR). The CAR refers to the zone at the junction of the TB
airways and the gas exchange region. This area is also termed the proximal alveolar
region (PAR) and is defined as the first generation distal to the terminal bronchioles.
Contained within the CAR, the respiratory bronchioles were confirmed as the site
receiving the greatest O3 dose (18O mass/lung weight) in resting O3 exposed rhesus
monkeys, when not considering the nose (Plopper et al., 1998). Furthermore, the
greatest cellular injury occurred in the vicinity of the respiratory bronchioles and was
dependent on the delivered O3 dose to these tissues (see also Section 5.4.1).
However, 18O label was detected to a lesser extent in other regions of the TB airway
tree, showing that O3 is delivered to these compartments as well, although in a
smaller dose. These studies agree with earlier model predictions showing that the
tissue O3 dose (O3 flux to liquid-tissue interface) was low in the trachea, increased to
a maximum in the terminal bronchioles and the CAR, and then rapidly decreased in
the alveolar region (Miller et al., 1985). It was also predicted that the net O3 dose (O3
flux to air-liquid interface) gradually decreased with distal progression from the
trachea to the end of the TB region and then rapidly decreased in the alveolar region.
Despite the exclusion of the URT and appreciable O3 reactions with ELF
constituents after the 16th generation, the results from the model agree with
experimental results showing that the greatest O3 tissue dose was received in the
CAR (Miller et al.. 1985).
5-9
-------
Inhomogeneity in the RT structure may affect the dose delivered to this target site.
Models have predicted that the farther the PAR is from the trachea, the less the O3
tissue dose to the region. Ultman and Anjilvel (1990) and Overton and Graham
(1989) predicted approximately a 50 to 300% greater PAR dose for the shortest path
relative to the longest path in humans and rats, respectively. In addition, Mercer et al.
(1991) found that both path distance and ventilatory unit size affected dose.
The variation of O3 dose among anatomically equivalent ventilatory units was
predicted to vary as much as 6-fold, as a function of path length from the trachea.
This could have implications in regional damage to the LRT, such that even though
the average LRT dose may be at a level where health effects would not be predicted,
local regions of the RT may receive considerably higher than average doses and
therefore be at greater risk of effects.
Since the URT is the first part of the RT to be exposed to O3, the nasal membranes
are another target site at risk of injury from inhaled O3. Injury to the nasal epithelium
has been shown to be site-specific (see Section 5.3.7) and studies have found that the
location of reactive gas-induced nasal lesions may be attributable to the local dose of
gas reaching that area (Garcia et al., 2009a; Morgan and Monticello, 1990). Similar
to the LRT, inhomogeneity of the nasal anatomy, nasal fluid composition, and
ventilation and airflow patterns affects the uptake of O3 into the nasal passageways.
5.2.2.3 Upper Respiratory Tract Ozone Removal and Dose
Total O3 uptake in the entire RT in rats and guinea pigs is 40-54% efficient (Hatch et
al.. 1989: Wiester et al.. 1988: Wiester et al.. 1987). while in humans at rest it ranges
from 80-95% efficient (Huetal. 1992). The URT provides a defense against O3
entering the lungs by removing half of the O3 that will be absorbed from the
airstream. In both animals and humans, about 50% of the O3 that was absorbed in the
RT was removed in the head (nose, mouth, and pharynx), about 7% in the
larynx/trachea, and about 43% in the lungs (Huetal., 1992: Hatch et al., 1989: Miller
et al., 1979). However, experimental studies in dogs have reported 75-100% uptake
in the URT (Yokoyama and Frank, 1972: Vaughan et al., 1969). The fraction of O3
taken up was inversely related to flow rate and to inlet O3 concentration (Yokoyama
and Frank, 1972: Vaughan et al., 1969). URT absorption is relatively high due in part
to the large surface area of the nasal airways. The limiting factors in nasal O3 uptake
were simultaneous diffusion and chemical reaction of O3 in the nasal ELF (Santiago
et al., 2001). The ELF layer in the nose is thicker than in the rest of the RT, and
mathematical estimates predicted that O3 penetrates less than the thickness of the
ELF layer; reaction products are likely the agents damaging the nasal tissue and not
O3 itself. It was hypothesized that the nasal non-linear kinetics of O3 uptake fraction
result from the depleting substrates in the nasal ELF becoming the limiting factor of
the reaction (Santiago et al., 2001).
Uptake efficiencies have been measured for various segments of the URT
(Table 5-1). Gerrity et al. (1995) reported unidirectional uptake efficiencies of O3
5-10
-------
inhaled from a mouthpiece; of 0.18 from the mouth to vocal cords, 0.095 from the
vocal cords to the upper trachea (totaling 0.27), 0.084 from the upper trachea to the
main bifurcation carina (total cumulative efficiency from the mouth of 0.36), and
essentially zero between the carina and the bronchus intermedius (total cumulative
efficiency from the mouth of 0.33). These values are lower than those calculated by
Hu et al. (1992) that reported cumulative uptake efficiencies of 0.21, 0.36, 0.44, and
0.46 during a complete breath in which an O3 bolus penetrated between the mouth
and the vocal cords, the upper trachea, the main bifurcation carina, and the bronchus
intermedius, respectively. The lower efficiencies seen in Gerritv et al. (1995) may
have resulted because these investigators measurements were based on inhalation
alone or were caused by O3 scrubbing by the mouthpiece.
Past studies investigating nasal uptake of O3 have shown that the nose partially
protects the LRT from damage from inspired O3 (Santiago et al., 2001; Gerritv et al.,
1988). Sawyer et al. (2007) further investigated nasal uptake of O3 in healthy adults
during exercise. Fractional O3 uptake, acoustic rhinometry (AR), and nasal NO
measurements were taken in ten adults (8 women, 2 men) exposed to 200 ppb O3
before and after moderate exercise at two flow rates (10 and 20 L/min). The percent
nasal uptake of O3 was -50% greater at 10 L/min compared to 20 L/min both pre-
and post-exercise. However, the inhaled O3 dose delivered to the LRT (i.e., flow rate
x exposure concentration x (1 - nasal absorbed fraction)) was 1.6-fold greater at the
higher flow than at the lower flow (2.5 compared to 0.9 ppm-L/min). Prior exercise
did not affect O3 uptake at either flow rate, but did significantly increase nasal
volume (Vn) and AR measurements of nasal cross-sectional area (minimum cross-
sectional area (MCA) that corresponds to the nasal valve, CSA2 that corresponds to
the anterior edge of the nasal turbinates, and CSA3 that corresponds to the posterior
edge of the nasal turbinates) (p < 0.05) (Sawyer et al.. 2007). Conversely, exercise
decreased nasal resistance (Rn) (p <0.01) and NO production (nonsignificant,
p >0.05). The change in Vn and CSA2:MCA ratio was correlated with the percent
change in nasal uptake, however the overall effect was small and sensitive to
elimination of outliers and sex segregation.
Overall, the majority of studies suggest that the URT removes about half of the O3
that will be absorbed by reactions in the nasal ELF. The exact uptake efficiency is
dependent on variations in flow rate and inhaled concentration.
5.2.2.4 Lower Respiratory Tract Ozone Uptake and Dose
Approximately 43% of the O3 absorption occurs in the LRT of both humans and
animals. Models predicted that the net O3 dose decreases distally from the trachea
toward the end of the TB region and then rapidly decreases in the alveolar region
(Miller et al.. 1985). Further, these models predicted low tissue O3 dose in the
trachea and large bronchi.
Uptake efficiency depends on a number of variables, including O3 exposure
concentration, exposure time, and breathing pattern. For breaths of similar
5-11
-------
waveforms, respiratory patterns are uniquely described by breathing frequency (fB)
and tidal volume (VT); by minute ventilation (VE = fB x VT) and fB; or by VE and VT.
Simulations from the Overton et al. (1996) single-path anatomical respiratory tract
model, where the upper and lower respiratory tracts were modeled but uptake by the
URT was not considered, predicted that fractional uptake and PAR O3 dose increased
with VT when fB was held constant. Likewise, experimental studies found that O3
uptake was positively correlated with changes in VT (TJltman et al., 2004; Gerrity et
al., 1988). Also, O3 exposure led to a reflex mediated increase in fB and reduction in
VT, hypothesized to be protective by decreasing the dose delivered to the lung at a
particular VE (Gerrity et al., 1994). Nasal O3 uptake efficiency was inversely
proportional to flow rate (Santiago et al., 2001), so that an increase in VE will
increase O3 delivery to the lower airways. At a fixed VE, increasing VT
(corresponding to decreasing fB) drove O3 deeper into the lungs and increased total
respiratory uptake efficiency (Figure 5-4) (Ultman et al., 2004; Wiester et al., 1996a;
Gerrity et al., 1988). Modeling predicted a decrease in fractional uptake with
increased fB when VT was held constant, but an increase in PAR dose with increased
fB (Overton et al., 1996). Similarly, increased fB (80 - 160 bpm) and shallow
breathing in rats decreased midlevel tracheal 18O content and an increased 18O
content in the mainstem bronchi (Alfaro et al., 2004). This dependence may be a
result of frequency-induced alterations in contact time that affects the first-order
absorption rate for O3 (Postlethwait et al., 1994). Also, an association of O3 uptake
efficiency was found with VE and exposure time.
Increasing flow leads to deeper penetration of O3 into the lung, such that a smaller
fraction of O3 is absorbed in the URT and uptake shifts to the TB airways and
respiratory airspaces (Nodelman and Ultman. 1999; Hu et al.. 1994; Ultman et al..
1994). Huetal. (1994) and Ultman et al. (1994) found that O3 absorption increased
with volumetric penetration (Vp) of a bolus of O3 into the RT. Ozone uptake
efficiency and Vp were not affected by bolus O3 concentration (Kabel et al.. 1994;
Hu et al.. 1992). indicating that under these experimental conditions O3 uptake was a
linear absorption process, where the diffusion and chemical reaction rates of O3 were
proportional to the O3 concentration. The absorption relationship would not be linear
once interfacial mass transfer was saturated. As mentioned above, a weak negative
relationship between O3 concentration and uptake efficiency was reported for the
nasal cavities by Santiago et al. (2001). Rigas et al. (2000) also found a weak but
significant negative dependence of O3 concentration on RT uptake efficiency in
exercising individuals. This study also found that exposure time had a small but
significant influence on uptake efficiency; however, this negative dependence may be
an artifact of progressive depletion of reactive substrates from the ELF.
5-12
-------
n.u -
Ł 0.9 -
o>
jo
0>
JS 0.8 -
a
3
0.7 -
+ ++ + + o
. +
-H-
*°+ °/ ° 0
+ e ++<
art- o
o o
0
0
O 28 Women
+ 32 Men 0
i i i i
20 30 40 50
Breathing Frequency (bpm)
Note: Subjects breathed 250 ppb O3 oronasally via a breathing mask. The uptake efficiency was well correlated with breathing
frequency (r = -0.723, p <0.001) and tidal volume (not illustrated; r = 0.490, p <0.001).
Source: Reprinted with permission of Health Effects Institute (Ultman et al.. 2004).
Figure 5-4 Total O3 uptake efficiency as a function of breathing frequency at a
constant minute ventilation of 30 L/min.
Past studies have shown that O3-induced epithelial damage to the lung occurs with a
reproducible pattern of severity between daughter branches of individual bifurcations
that is dependent on the O3 concentration-time profile of the inhaled gas.
A 3-D computational fluid dynamics model was created to investigate the O3
transport in a single airway bifurcation (Taylor et al.. 2007). The model consisted of
one parent branch and two symmetrical daughter branches with a branching angle of
90° and a sharp carinal ridge. Various flow scenarios were simulated using Reynolds
numbers (Re) ranging from 100 to 500. The Reynolds number that corresponds to a
certain airway generation is dependent upon both lung size and VE, such that the
range in Reynolds numbers, from 100-500, would encompass generations 1-5, 3-7,
and 6-10 for an adult during quiet breathing, light exertion, and heavy exercise,
respectively, whereas the same Reynolds number range corresponds to generations
0-4, 1-6, and 4-8 for a 4-year-old child. This model predicted velocity distributions
that were consistent with earlier work of Schroter and Sudlow (1969). and also
reported O3 concentration and wall uptake distributions. The model predicted that
during inspiration, the velocity and O3 concentration distribution were axisymmetric
throughout the parent branch, but skewed toward the inner wall within the daughter
branches. During expiration, the model predicted that the velocity and O3
concentration distribution was slightly skewed toward the outer walls of the daughter
branches. Hot spots of wall flux existed at the carina during inspiration and
5-13
-------
expiration with Re >100. Additional hot spots were found during expiration on the
parent branch wall downstream of the branching region.
Overall O3 inhalation uptake in humans is over 80% efficient, but the exact
efficiency that determines how much O3 is available at longitudinally distributed
compartments in the lung is sensitive to changes in VT, fe, and to a minor extent,
exposure time.
5.2.2.5 Mode of Breathing
Ozone uptake and distribution is sensitive to the mode of breathing. Variability in TB
airways volume had a weaker influence on O3 absorption during nasal breathing
compared to oral breathing. This could be a result of O3 scrubbing in the nasal
passageways that are bypassed by oral breathing. Studies by Ultman and colleagues,
using bolus inhalation in humans, demonstrated that O3 uptake fraction into the
upper airways was greater during nasal breathing than during oral breathing
(e.g., 0.90 during nasal breathing and 0.80 during oral breathing at 150 mL/sec and
0.45 during nasal breathing and 0.25 during oral breathing at 1,000 mL/sec)
(Nodelman and Ultman. 1999: Kabel et al. 1994: Ultman et al. 1994). Therefore,
oral breathing results in deeper penetration of O3 into the RT with a higher absorbed
fraction in the TB and alveolar airways (Nodelman and Ultman. 1999). Similar
results were also obtained from O3 uptake studies in dogs (Yokoyama and Frank.
1972). Earlier human studies suggested that oral or oronasal breathing results in a
higher O3 uptake efficiency than nasal breathing (Wiester et al.. 1996a: Gerrity et al..
1988). Overall, the mode of breathing may have a seemingly small effect on the RT
uptake efficiency; however, it does play an important role in the distribution of O3
deposited in the distal airways.
5.2.2.6 Interindividual Variability in Dose
Similarly exposed individuals vary in the amount of actual dose delivered to the LRT
(Santiago et al.. 2001: Rigas et al.. 2000: Bushetal.. 1996). Interindividual
variability accounted for between 10-50% of the absolute variability in O3 uptake
measurements (Santiago et al.. 2001: Rigas et al.. 2000). When concentration, time,
and VE were held constant, fractional absorption ranged from 0.80 to 0.91 (Rigas et
al.. 2000). It has been hypothesized that interindividual variation in O3 induced
responses such as FEVi is the result of interindividual variation in net dose or
regional O3 uptake among exposed individuals.
Recent studies have reiterated the importance of intersubject variation in O3 uptake.
The intersubject variability in nasal O3 uptake determined by Sawyer et al. (2007)
ranged from 26.8 to 65.4% (pre- and post-exercise). A second study investigating the
use of the CO2 expirogram to quantify pulmonary responses to O3 found that
5-14
-------
intersubject variability accounted for 50% of the overall variance in the study (Taylor
et al.. 2006).
Variability in net or tissue dose may be attributed to differences in the pulmonary
physiology, anatomy, and biochemistry. Since the URT and TB airways remove the
majority of inhaled O3 before it reaches the gas exchange region, the volume and
surface area of these airways will influence O3 uptake. Models predicted that
fractional O3 uptake and PAR dose (flux of O3 to the PAR surfaces divided by
exposure concentration) increase with decreasing TB volume and decreasing TB
region expansion. On the contrary, alveolar expansion had minimal effect on uptake
efficiency as relatively little O3 reaches the peripheral lung (Bush et al., 2001;
Overton et al., 1996). Ozone uptake was virtually complete by the time O3 reaches
the alveolar spaces of the lung (Postlethwait et al., 1994). Experimental studies have
found that differences in TB volumes may account for 75% of the variation in
absorption between subjects (Ultman et al., 2004). In support of this concept,
regression analysis showed that O3 absorption was positively correlated with
anatomical dead space (VD) and TB volume (i.e., VD minus VURT), but not total lung
capacity (TLC), forced vital capacity (FVC), or functional residual capacity (FRC)
(Ultman et al.. 2004: Bushetal.. 1996: Huetal.. 1994: Postlethwait et al.. 1994).
Variability in VD was correlated more with the variability in the TB volume than the
URT volume. Similarly, uptake was correlated with changes in individual bronchial
cross-sectional area, indicating that changes in cross-sectional area available for gas
diffusion are related to overall O3 retention (Reeser et al., 2005: Ultman et al., 2004).
When coupled, these results suggest that the larger surface-to-volume ratio
associated with the smaller airways in women enhances local O3 uptake, thereby
reducing the distal penetration volume of O3 into the female respiratory system.
When absorption data were normalized to Vp/VD, variability attributed to sex
differences were not distinguishable (Bush et al.. 1996). These studies provide
support to the RT anatomy, especially the TB volume and surface area, playing a key
role in variability of O3 uptake between individuals.
In addition, variability between individuals is influenced by age. Overton and
Graham (1989) predicted that the total mass of O3 absorbed per minute (in units of:
|_ig/min per [|_ig/m3 of ambient O3]) increased with age from birth to adulthood. This
model predicted that during quiet breathing the LRT distribution of absorbed O3 and
the CAR O3 tissue dose were not sensitive to age. However, during heavy exercise or
work O3 uptake was dependent on age. A physiologically based pharmacokinetic
model simulating O3 uptake predicted that regional extraction of O3 was relatively
insensitive to age, but extraction per unit surface area was 2-fold to 8-fold higher in
infants compared to adults, due to the fact that children under age 5 have much a
much smaller airway surface area in the extrathoracic (nasal) and alveolar regions
(Sarangapani et al., 2003). Additionally, children tend to have a greater oral
breathing contribution than adults at rest and during exercise (Bennett et al., 2008:
Becquemin et al., 1999: James et al., 1997). Even after adjusting for differences in
surface area, the dose rate to the lower airways of children compared to adults is
increased further because children breathe at higher minute ventilations relative to
their lung volumes.
5-15
-------
Smoking history, with its known increase in mucus production, was not found to
affect the fractional uptake of a bolus of O3 in apparently healthy smokers with
limited smoking history (Bates et al., 2009). Despite similar internal O3 dose
distribution, the smokers exhibited greater pulmonary responses to O3 bolus
exposures, measured as FEVi decrements and increases in the normalized slope of
the alveolar plateau (SN). This was contrary to previous studies conducted in smokers
with a greater smoking history that found decreased O3 induced decrements in FEVi
in smokers during continuous O3 exposure (Frampton et al.. 1997a: Emmons and
Foster. 1991).
5.2.2.7 Physical Activity
Exercise increases the overall exposure of the lung to inhaled contaminants due, in
most part, to the increased intake of air. Thus, human studies have used exercise, at a
variety of activity levels, to enhance the effects of O3 (Table 5-2). Further
explanation of the effects of physical activity on ventilation can be found in Chapters
4 and 6. Table 4-5 presents the mean ventilation rates at different activity levels for
different age groups. Table 6-1 provides activity levels as detailed in specific human
exposure studies.
Table 5-2 General adult human inhalation rates by activity levels.
Activity Level Inhalation Rate
Light 2 to 3 x resting VE a
Moderate 4 to 6 x resting VE
Heavy 7 to 8 x resting VE
Very Heavy >9 x resting VE
"Resting VE approximates 8 L/min
Source: U.S. EPA (1986).
As exercise increases from a light to moderate level, VT increases. This increase in
VT is achieved by encroaching upon both the inspiratory and expiratory reserve
volumes of the lung (Dempsev et al.. 1990). After VT reaches about 50% of the vital
capacity, generally during heavy exercise, further increases in ventilation are
achieved by increasing fB. Ventilatory demands of very heavy exercise require
airway flow rates that often exceed 10 times resting levels and VT that approach 5
times resting levels (Dempsev et al.. 2008).
In addition to increasing the bulk transport of O3 into the lung, exercise also leads to
a switch from nasal to oronasal breathing. Higher ventilatory demand necessitates a
lower-resistance path through the mouth. The contribution of nasal breathing to the
VE varies as a function of age, sex, and race. Children tended to have a lesser nasal
contribution to breathing than adults at rest and during exercise at matched percent
5-16
-------
maximum work (Bennett et al., 2008). Males had less nasal contribution to breathing
at rest and during exercise at matched percent maximum work compared with
females (Bennett et al., 2003). The difference between the sexes may be explained by
the difference in VE at a given percent maximal workload. Females had a lower VE
than males so had to augment breathing orally at higher work efforts. Caucasians had
a lesser nasal contribution than African-Americans at rest and during exercise at
matched percent maximum work (Bennett et al., 2003).
This increase in VT and flow associated with exercise in humans shifts the net O3
dose further into the periphery of the RT causing a disproportionate increase in distal
lung tissue dose. Modeling heavy exercise by increasing ventilatory parameters from
normal respiration levels predicted a 10-fold increase in total mass uptake of O3
(Miller et al., 1985). This model also predicted that as exercise and ventilatory
demand increased, the maximum tissue dose, the O3 reaching the tissues, moved
distally into the RT (Figure 5-5). When the flow was increased to what is common in
moderate or heavy exercise (respiratory flow = 45-60 L/min compared to 15 L/min),
the URT absorbed a smaller fraction of the O3 (0.10 at high flow rate to -0.50 at low
flow rate); however, the trachea and more distal TB airways received higher doses
during higher flow rates than at lower flow rates (0.65 absorbed in the lower TB
airways, and 0.25 absorbed in the alveolar zone with high flow compared to 0.5 in
the TB with almost no O3 reaching the alveolar zone at low flow) (Hu et al., 1994).
The same shift in the O3 dose distribution more distally in the lung occurred in other
studies mimicking the effects of exercise (Nodelman and Ultman, 1999). Also, LRT
uptake efficiency was sensitive to age only under exercise conditions (Overton and
Graham, 1989). The total mass of O3 absorbed per minute (|_ig/min per [|^g/m3 of
ambient O3]) was predicted to increase with age during heavy work or exercise.
A recent study by Sawyer et al. (2007) approximated that doubling minute
ventilation led to only a 1.6-fold higher delivered dose rate of O3 to the lung
(delivered dose was calculated as: flow rate x [O3 ppm] x (100-percent nasal O3
uptake)) due to a decrease in URT uptake with increasing flow rate. Past models
have predicted the increase in uptake during exercise is distributed unevenly in the
RT compartments and regions. Tissue and net dose in the TB region increased
~1.4-fold during heavy exercise compared to resting conditions, whereas the alveolar
surface layer and tissue uptake increased by factors of 5.2 and 13.6, respectively
(Miller et al.. 1985).
5-17
-------
10-*
4 8 12 16 20
AIRWAY GENERATION (Z)
-TB-
Note: Curve 1: VT = 500 ml; fB = 15 breaths/min. Curve 2: VT = 1,000 ml; fB = 15 breaths/min. Curve 3: VT = 1,750 ml; fB = 20.3
breaths/min. Curve 4: VT = 2,250 ml; fB = 30 breaths/min. TB = tracheobronchial region; P = pulmonary region.
Source: Reprinted with permission of Elsevier (Miller et al.. 1985).
Figure 5-5 Modeled effect of exercise on tissue dose of the LRT.
5.2.2.8 Summary
In summary, O3 uptake is affected by complex interactions between a number of
factors including RT morphology, breathing route, frequency, and volume,
physicochemical properties of the gas, physical processes of gas transport, as well as
the physical and chemical properties of the ELF and tissue layers. The role of these
processes varies throughout the length of the RT and as O3 moves from the gas into
liquid compartments of the RT.
About half of the O3 that will be absorbed from the airstream is removed in the URT,
which provides a defense against O3 entering the lungs. However, the local dose to
the URT tissue is site-specific and dependent on the nasal anatomy, nasal fluid
composition, and ventilation and airflow patterns of the nasal passageways.
5-18
-------
The primary uptake site of O3 delivery to the LRT epithelium is believed to be the
CAR, however, similar to the URT, inhomogeneity in the RT structure may affect the
dose delivered to this target site with larger path lengths leading to smaller locally
delivered doses. This could have implications in regional damage to the RT, such
that even though the average RT dose may be at a level where health effects would
not be predicted, local regions of the RT may receive considerably higher than
average doses and therefore be at greater risk of effects. Recent studies have
provided evidence for hot spots of O3 flux around bifurcations in airways.
Experimental studies and models have suggested that the net O3 dose gradually
decreases distally from the trachea toward the end of the TB region and then rapidly
decreases in the alveolar region. However, the tissue O3 dose is low in the trachea,
increases to a maximum in the terminal bronchioles and the CAR, and then rapidly
decreases distally into the alveolar region.
Ozone uptake efficiency is sensitive to a number of factors. Fractional absorption
will decrease with increased flow and increase proportional to VT, so that at a fixed
VE, increasing VT (or decreasing fB) drives O3 deeper into the lungs and increases
total respiratory uptake efficiency. Individual total airway O3 uptake efficiency is
also sensitive to large changes in O3 concentration, exposure time, and VE. Major
sources of variability in absorption of O3 include O3 concentration, exposure time,
fB, VE, and VT, but the interindividual variation is the greatest source of variability
uptake efficiency. The majority of this interindividual variability is due to differences
in TB volume and surface area.
An increase in VT and fB are both associated with increased physical activity. These
changes and a switch to oronasal breathing during exercise results in deeper
penetration of O3 into the lung with a higher absorbed fraction in the ET, TB, and
alveolar airways. For these reasons, increased physical activity acts to move the
maximum tissue dose of O3 distally into the RT and into the alveolar region.
5.2.3 Ozone Reactions and Reaction Products
Ozone dose is affected by the chemical reactions or the products of these reactions
that result from O3 exposure. The process by which O3 moves from the airway
lumen into the ELF is related to the coupled diffusion and chemical reactions
occurring in ELF and is called "reactive absorption." Ozone is chemically reactive
with a wide spectrum of biomolecules and numerous studies have evaluated the loss
of specific molecules such as GSH and the appearance of plausible products such as
nonanal. Both in vitro and in vivo studies contribute to the understanding of O3
reactions and reaction products.
Ozone may interact with many of the components in the ELF including
phospholipids, neutral lipids like cholesterol, free fatty acids, proteins, and low
molecular weight antioxidants as has been demonstrated in in vitro studies (Perez-
Gil, 2008; Uppuetal, 1995). It was estimated that 88% of the O3 that does not come
in contact with antioxidants will react with unsaturated fatty acids in the ELF
5-19
-------
including those contained within phospholipids or neutral lipids (Uppu et al., 1995).
Ozone reacts with the double bond of unsaturated fatty acids to form stable and less
reactive ozonide, aldehyde, and hydroperoxide reaction products via chemical
reactions such as the Criegee ozonolysis mechanism (Figure 5-6) (Prvor et al., 1991).
Lipid ozonation products, such as the aldehydes hexanal, heptanal, and nonanal, have
been recovered after O3 exposure in human BAL fluid (BALF), rat BALF, isolated
rat lung, and in vitro systems (Frampton et al.. 1999: Postlethwait et al.. 1998: Pryor
et al.. 1996). Adducts of the aldehyde 4-hydroxynonenal were found in human
alveolar macrophages after O3 exposure in vivo (Hamilton et al.. 1998).
Polyunsaturated fatty acid (PUFA) reactions are limited by the availability of O3
since lipids are so abundant in the ELF. Yields of O3-induced aldehydes were
increased by the decrease in other substrates such as ascorbic acid (AH2)
(Postlethwait et al.. 1998). Free radicals are also generated during O3-mediated
oxidation reactions with PUFA (Prvor. 1994). These reactions are reduced by the
presence of the lipid-soluble free radical scavenger a-tocopherol (a-TOH) (Prvor.
1994: Fujita et al.. 1987: Pryor. 1976). PUFA reactions may not generate sufficient
bioactive materials to account for acute cell injury, however only modest amounts of
products may be necessary to induce cytotoxicity (Postlethwait and Ultman. 2001:
Postlethwait et al.. 1998).
_ I I _ _
RHC — CH + Oo ^ RHC — CH— ^ RHC — O — O "^ RHC — O
PUFA ozone trioxo/ane carbonyl oxide aldehyde
either in /°~°x or in the /OH
the > RHC CH—presence + RHC >• RHC = O + H2O2
absence \n/' ofHoO V^u
of HoO u UUn aldehyde hydrogen
Criegee ozonide hydroxyhydroperoxy cpd. peroxide
Note: Not all secondary reaction products are shown.
Source: U.S. EPA (2006b).
Figure 5-6 Schematic overview of O3 interaction with PUFA in ELF and lung
cells.
Cholesterol is the most abundant neutral lipid in human ELF. Reaction of cholesterol
with O3 results in biologically active cholesterol products such as the oxysterols,
(3-epoxide and 6-oxo-3,5-diol (Murphy and Johnson. 2008: Pulfer et al.. 2005: Pulfer
and Murphy. 2004). Product yields depend on ozonolysis conditions, however
cholesterol ozonolysis products form in similar abundance to phospholipid-derived
ozonolysis products in rat ELF (Pulfer and Murphy. 2004).
5-20
-------
The ELF also contains proteins derived from blood plasma as well as proteins
secreted by surface epithelial cells. Ozone reactions with proteins have been studied
by their in vitro reactions as well as reactions of their constituent amino acids (the
most reactive of which are cysteine, histidine, methionine, tyrosine, and tryptophan).
Ozone preferentially reacts with biomolecules in the following order: thiosulfate
>ascorbate >cysteine ~ methionine >glutathione (Kanofsky and Sima. 1995). Rate
constants for the reaction of amino acids with O3 vary between studies due to
differing reaction conditions and assumptions; however aliphatic amino acids were
consistently very slow to react with O3 (e.g., alanine: 25-100 moles/L/sec) (Kanofsky
and Sima. 1995: Ignatenko and Cherenkevich. 1985: Pryor et al.. 1984: Hoigne and
Bader. 1983). Uppu et al. (1995) predicted that 12% of inhaled O3 that does not react
with antioxidants will react with proteins in the ELF.
Reactions of O3 with low molecular weight antioxidants have been extensively
studied. The consumption of antioxidants such as uric acid (UA), ascorbate (AH2),
and reduced glutathione (GSH) by O3 was linear with time and positively correlated
with initial substrate concentration and O3 concentration (Mudway and Kelly, 1998:
Mudway et al., 1996). Endogenous antioxidants are present in relatively high
concentrations in the ELF of the human airways (obtained as B ALF) and display
high (but not equal) intrinsic reactivities toward O3. In individual and in limited
composite mixtures, UA was the most reactive antioxidant tested, followed by AH2
(Mudway and Kelly, 1998). GSH was consistently less reactive than UA or AH2
(Mudwav and Kelly. 1998: Mudwav et al.. 1996: Kanofsky and Sima. 1995).
To quantify these reactions, Kermani et al. (2006) evaluated the interfacial exposure
of aqueous solutions of UA, AH2, and GSH (50-200 jJVl) with O3 (1-5 ppm). Similar
to the results of Mudwav and Kelly (1998), this study found the hierarchy in
reactivity between O3 and these antioxidants to be UA> AH2»GSH. UA and AH2
shared a 1:1 stoichiometry with O3, whereas 2.5 moles of GSH were consumed per
mole of O3. Using these stoichiometries, reaction rate constants were derived
(5.8xl04M-1sec"1, S.SxlO4]^1 sec"1, and 57.5 M'0'75 sec"1 [20.9 M"1 sec"1] for the
reaction of O3 with UA, AH2, and GSH, respectively). Other studies report reactive
rate constants that are two to three orders of magnitude larger, however these studies
used higher concentrations of O3 and antioxidants under less physiologically relevant
experimental conditions (Kanofsky and Sima. 1995: Giamalva et al.. 1985: Pryor et
al.. 1984). Since O3 acts through competition kinetics, the effective concentration of
the reactants present in the ELF will determine the reactions that occur in vivo. For
example, the pKa of GSH is about 8.7 so that at physiological pH very little is in the
reactive form of thiolate (GS~). On the other hand, ascorbic acid has a pKa of about
4.2 so it exists almost entirely as ascorbate (AH") in the ELF. Thus, the effective
concentration of GSH that is available to react with O3 will be much lower than that
of ascorbate in ELF.
A series of studies used new techniques to investigate the reaction products resulting
from initial air-liquid interface interactions of O3 with ELF components
(e.g., antioxidants and proteins) in ~1 millisecond (Enami et al., 2009a, b, c, 2008a,
b). Solutions of aqueous UA, AH2, GSH, a-TOH, and protein cysteines (CyS) were
sprayed as microdroplets in O3/N2 mixtures at atmospheric pressure and analyzed by
5-21
-------
electrospray mass spectrometry. These recent studies in which the large surface to
volume ratio of microdroplets promote an interfacial reaction demonstrated different
reactivity toward AH2, UA, and GSH by O3 compared to previous studies using bulk
liquid phase bioreactors. This artificial system does not recapitulate the lung surface
so caution must be taken in translating the results of these studies to in vivo
conditions.
As was seen in previous studies (Kermani et al.. 2006; Kanofsky and Sima, 1995),
the hierarchy of reactivity of these ELF components with O3 was determined to be
AH2 ~ UA > CyS >GSH. There was some variance between the reaction rates and
product formation of UA, AH2, and GSH with O3 as investigated by Enami et al.
(2009a, b, c, 2008a, b) versus O3 reacting with bulk liquid phase bioreactors as
described previously. UA was more reactive than AH2 toward O3 in previous studies,
but in reactions with O3 using microdroplets, these antioxidants had equivalent
reactivity (Enami et al., 2008b). As O3 is a kinetically slow one-electron acceptor but
very reactive O-atom donor, products of the interaction of O3 with UA, AH2, GSH,
CyS, and a-TOH result from addition of n O-atoms (n = 1-4). These products
included epoxides (e.g., U-O"), peroxides (e.g., U-O2"), and ozonides (e.g., U-O3").
For instance, GSH was oxidized to sulfonates (GSO3~/GSO32~), not glutathione
disulfide (GSSG) by O3 (Enami et al., 2009b). However, it is possible that other
oxidative species are oxidizing GSH in vivo, since sulfonates are not detected in O3
exposed ELF whereas GSSG is. This is also supported by the fact that O3 is much
less reactive with GSH than other antioxidants, such that <3% of O3 will be
scavenged by GSH when in equimolar amounts with AH2 (Enami et al., 2009b).
This series of studies also demonstrated that ozonolysis product yields and formation
were affected by pH. Acidified conditions (pH ~ 3-4), such as those that may result
from acidic particulate exposure or pathological conditions like asthma (pH ~ 6),
decreased the scavenging ability of UA and GSH for O3; such that at low pH, the
scavenging of O3 must be taken over by other antioxidants, such as AH2 (Enami et
al., 2009b, 2008b). Also, under acidic conditions (pH ~ 5), the ozonolysis products of
AH2 shifted from the innocuous dehydro-ascorbic acid to the more persistent
products, AH2 ozonide and threonic acid (Enami et al., 2008a). It is possible that the
acidification of the ELF by acidic copollutant exposure will increase the toxicity of
O3 by preventing some antioxidant reactions and shifting the reaction products to
more persistent compounds.
Since ELF exists as a complex mixture, it is important to look at O3 reactivity in
substrate mixtures. Individual antioxidant consumption rates decreased as the
substrate mixture complexity increased (e.g., antioxidant mixtures and albumin
addition) (Mudway and Kelly. 1998). However, O3 reactions with AH2
predominated over the reaction with lipids, when exposed to substrate solution
mixtures (Postlethwait et al.. 1998). It was suggested that O3 may react with other
substrates once AH2 concentrations within the reaction plane fall sufficiently.
Additionally, once AH2 was consumed, the absorption efficiency diminished,
allowing inhaled O3 to be distributed to more distal airways (Postlethwait et al..
1998). Multiple studies have concluded O3 is more reactive with AH2 and UA than
5-22
-------
with the weakly reacting GSH (or cysteine or methionine) or with amino acid
residues and protein thiols (Kanofsky and Sima, 1995; Cross et al., 1992).
In a red blood cell (RBC) based system, AH2 augmented the in vitro uptake of O3 by
6-fold, as computed by the mass balance across the exposure chamber (Ballinger et
al.. 2005). However, estimated in vitro O3 uptake was not proportional to the
production of O3-derived aldehydes from exposing O3 to RBC membranes (Ballinger
et al., 2005). In addition, O3 induced cell membrane oxidation that required
interactions with AH2 and GSH, but not UA or the vitamin E analog Trolox. Further,
aqueous phase reactions between O3 and bovine serum albumin did not result in
membrane oxidation (Ballinger et al., 2005). The presence of UA or bovine serum
albumin protected against lipid and protein oxidation resulting from the reaction of
O3 and AH2 (Ballinger et al., 2005). This study provided evidence that antioxidants
may paradoxically facilitate O3-mediated damage. This apparent contradiction
should be viewed in terms of the concentration-dependent role of the ELF
antioxidants. Reactions between O3 and antioxidant species exhibited a biphasic
concentration response, with oxidation of protein and lipid occurring at lower, but
not higher, concentrations of antioxidant. In this way, endogenous reactants led to the
formation of secondary oxidation products that were injurious and also led to
quenching reactions that were protective. Moreover, the formation of secondary
oxidation products mediated by some antioxidants was opposed by quenching
reactions involving other antioxidants.
Alterations in ELF composition can result in alterations in O3 uptake. Bolus O3
uptake in human subjects can be decreased by previous continuous O3 exposure
(120-360 ppb), possibly due to depletion of compounds able to react with O3 (Rigas
et al.. 1997: Asplund et al.. 1996). Conversely, O3 (360 ppb) bolus uptake was
increased with prior NO2 (360-720 ppb) or SO2 (360 ppb) exposure (Rigas et al.,
1997). It was hypothesized that this increased fractional absorption of O3 could be
due to increased production of reactive substrates in the ELF due to oxidant-induced
airway inflammation.
Besides AH2, GSH and UA, the ELF contains numerous antioxidant substances that
appear to be an important cellular defense against O3 including a-TOH, albumin,
ceruloplasmin, lactoferrin, mucins, and transferrin (Mudwav et al.. 2006: Freed et al..
1999). The level and type of antioxidant present in ELF varies between species,
regions of the RT, and can be altered by O3 exposure. Mechanisms underlying the
regional variability are not well-understood. It is thought that both plasma
ultrafiltrate and locally secreted substances contribute to the antioxidant content of
the ELF (Mudway et al.. 2006: Freed et al.. 1999). In the case of UA, the major
source appears to be the plasma (Peden et al.. 1995). Repletion of UA in nasal lavage
fluid was demonstrated during sequential nasal lavage in human subjects (Mudway et
al.. 1999a). When these subjects, exercising at a moderate level, were exposed to
200 ppb O3 for 2 hours, nasal lavage fluid UA was significantly decreased while
plasma UA levels were significantly increased (Mudway et al.. 1999a). The finding
that UA, but not AH2 or GSH, was depleted in nasal lavage fluid indicated that UA
was the predominant antioxidant with respect to O3 reactivity in the nasal cavity
5-23
-------
(Mudway et al., 1999a). However, in human BALF samples, the mean consumption
of AH2 was greater than UA (Mudway et al., 1996). In addition, concentrations of
UA were increased by cholinergic stimulation of the airways in human subjects,
which suggested that increased mucosal gland secretions were an important source
(Peden et al., 1993). Using the O3-specific antioxidant capacity assay on human nasal
lavage samples, Rutkowski et al. (2011) concluded that about 30% of the antioxidant
capacity of the nasal ELF was attributed to UA activity. Additionally, more than 50%
of the subject-to-subject differences in antioxidant capacity were driven by
differences in UA concentration. However, day-to-day within-subject variations in
measured antioxidant capacity were not related to the corresponding variations in UA
concentration in the nasal lavage fluid. Efforts to identify the predominant
antioxidant(s) in other RT regions besides the nasal cavity have failed to yield
definitive results.
Regulation of AH2, GSH and a-TOH concentrations within the ELF is less clear than
that of UA (Mudway et al., 2006). In a sequential nasal lavage study in humans,
wash-out of AH2 and GSH occurred, indicating the absence of rapidly acting
repletion mechanisms (Mudway et al., 1999a). Other studies demonstrated increases
in BALF GSH and decreases in BALF and plasma AH2 levels several hours
following O3 exposure (200 ppb for 2 h, while exercising at a moderate level)
(Mudwayetal.,2001; Blomberg et al., 1999; Mudway et al., 1999b). Studies with
rats exposed to 0.4-1.1 ppm O3 for 1-6 hours have shown consumption of AH2 that
correlates with O3 exposure (Gunnison and Hatch, 1999; Gunnison et al., 1996;
Vincent et al., 1996b). Further, cellular uptake of oxidized AH2 by several cell types
followed by intracellular reduction and export of reduced AH2 has been
demonstrated in vitro (Welch et al., 1995).
A body of evidence suggests that reaction of O3 within the ELF limits its diffusive
transport through the ELF; direct contact of O3 with the apical membranes of the
underlying epithelial cells therefore might be negligible in many regions of the RT
(Ballingeretal.,2005; Connor et al., 2004; Postlethwait and Ultman, 2001; Pryor,
1992). This conclusion is based on computational analyses and in vitro studies.
Direct confirmation using in vivo studies is limited. Nevertheless, when predicting
exposure-related outcomes across species and anatomic sites, whether O3 directly
contacts the apical membranes of the epithelial cells is an important consideration,
given that the extracellular surface milieu of the RT appreciably varies in terms of
the types and concentrations of the substrates present and the thickness of the ELF.
For O3 or its reaction products to gain access to the underlying cellular
compartments, O3 must diffuse at the air-liquid interface of the airway surface and
travel through the ELF layer. In vitro experiments have shown that O3 disappearance
from the gas phase depends on the characteristics of the ELF substrates (Postlethwait
et al.. 1998; Hu et al.. 1994). The ELF is comprised of the airway surface lining,
which includes the periciliary sol layer and overlying mucus gel layer, and the
alveolar surface lining, which includes the subphase of liquid and vesicular surfactant
and the continuous surfactant monolayer (Bastacky et al.. 1995). There is a
progressive decrease in ELF thickness and increase in interfacial surface with
5-24
-------
progression from the TB region to the alveolus (Mercer et al., 1992). The progressive
thinning of the ELF while moving further down the RT decreases the radial distance
O3 or its reaction products must travel to reach the cells lining the RT.
Taking into account the high reactivity and low water solubility of O3, Pry or (1992)
estimated the distance that O3 can penetrate into an ELF layer before it reacts with
endogenous substrates to form other more long-lived reactive species, thus initiating
a reaction cascade. These calculations utilize the Einstein-Smoluchowski equation to
compare the time (tdiff) for O3 to diffuse a distance (d) to the half-life (t^) of O3 in its
simultaneous reaction with substrates (Equation 5-1).
= d2/2D03 and trx = ln2/ksCs
Equation 5-1
where D03 is the O3 diffusion coefficient in ELF, ks is the bimolecular reaction rate
constant of O3 with a reactive substrate (s) in ELF, and Cs is the molar substrate
concentration. Importantly, it is assumed in the derivation of t^ that the substrate is
far in excess of O3 so that Cs is spatially uniform in the ELF. To within some
proportionality constant, the distance that O3 penetrates can be estimated by equating
to trx such that
d oc (D03/ksCs^/2
Equation 5-2
There is reasonable certainty that the O3 diffusion coefficient anywhere in the ELF is
in the range of D03~10"5 - 10"6 cm2/sec, but values of the ksCs product for the
reaction of O3 with specific substrates are much less reliable. Moreover, it is
unknown which substrates make the most important contributions to ksCs and how
these contributions vary from airway region to airway region. By asserting that
polyunsaturated fatty acids are the primary reactive substrate, Miller et al. (1985)
estimated that ksCs = 1,198 sec"1 in airway surface lining fluid and ksCs = 21.4 sec"1
in alveolar surface lining fluid. Pryor (1992) estimated the value of ksCs = 10"6 sec"1,
by assuming reduced glutathione is the primary substrate in airway surface lining
fluid. A value of ksCs =2.5 x 10s sec"1 was extracted from in vivo measurements of
O3 uptake into the airway surface lining fluid of the nasal cavities (Santiago et al.,
2001). These studies suggest that there is an uncertainty in the magnitude of ksCs
within airway surface lining fluid by a factor of 1,000, and that ksCs may be more
than 100 times greater in airway surface lining fluid than in alveolar surface lining
fluid.
With their estimates of ksCs = 106 sec"1 and D03 = 10"6 cm2/sec, Pryor(1992)
concluded that O3 could not penetrate an airway surface lining layer even as thin as
0.1 |^m. Comparable computations with the ksCs = 1,198 sec"1 value from Miller et
5-25
-------
al. (1985) would indicate that O3 penetrates an airway surface lining layer as thick as
3 |^m. Since airway surface lining layer thickness is on the order of 10 j^m in large
airways and 0.1 j^m in small airways, results using different estimates of ksCs have
entirely different implications regarding the direct role of O3 in damage to
underlying epithelium versus the role of toxic reaction products.
In the nasal passages, in particular, a diffusion analysis of in vivo O3 uptake
measurements made at different air flows indicated that the O3 penetration distance
(0.5 |^m) is considerably less than the thickness of the nasal surface lining layer
(10 |^m) (Santiago et al., 2001). A computational fluid dynamics model was able to
predict experimentally measured O3 uptake when the presences of a nasal surface
lining layer thickness was considered (Cohen-Hubal et al., 1996), further indicating
the need to properly account for the reaction-diffusion processes in the surface lining
layer.
Despite calculations and in vitro studies suggesting that reactions of O3 with
underlying epithelial cells may be negligible in some regions of the RT, there is some
evidence that suggests direct interaction of O3 with epithelial cells is possible. While
moving distally in the lung, the ELF thickness decreases and becomes ultra thin in
the alveolar region, possibly allowing for direct interaction of O3 with the underlying
epithelial cells. One definitive study conducted in excised rat lung measured alveolar
surface lining layer thickness over relatively flat portions of the alveolar wall to be
0.14 |^m, to be 0.89 |^m at the alveolar wall junctions, and 0.09 |^m over the
protruding features (Bastacky et al.. 1995). The area-weighted average thickness of
the alveolar surface lining fluid was found to be about 0.2 j^m and the alveolar
surface lining layer was continuous over the entire alveolar surface measured.
The surface appeared smooth; and no epithelial surface features or macrophage
features protruded above the air-liquid interface. It was noted that measurements of
alveolar surface lining layer thickness were made in lungs prepared in a state of
roughly 80% of total lung capacity, and as a result, the values reported would be
approaching the lowest values possible during the respiratory cycle. However, 4% of
the surface area in the alveolar compartment was covered by alveolar lining fluid
layer of less than 20 nm (Bastacky et al., 1995), suggesting the possibility that
unreacted O3 could penetrate to the cell layer in this region. Further it remains a
possibility that airways macrophages may protrude into the gas phase, allowing for
direct contact between O3 and airways epithelial cells.
Still, direct reaction of O3 with alveolar epithelial cells or macrophages may be
limited by the presence of dipalmitoyl phosphatidylcholine (DPPC), the major
component of surfactant, which has been shown in vitro to inhibit uptake of O3 into
an aqueous compartment containing ascorbate, glutathione, and uric acid (Connor et
al.. 2004). Further, the amount of O3 available to the alveolar compartment may be
limited by uptake of O3 in nasal and TB compartments. In fact, the amount of 18O
reaction product was lower in the alveolar tissues than in TB tissues of rhesus
monkeys immediately following a 2 hour exposure to 18O-labeled O3 (0.4 and 1 ppm)
(Plopper et al.. 1998). These considerations illustrate the difficulty in determining
whether O3 reacts directly with cells in the alveolar compartment.
5-26
-------
In some cases, however, with regard to the initiating mechanisms of cellular
perturbations, the precise reactive species that encounters the epithelia might or
might not have specificity to O3 per se or to its secondary oxidants. Many of the
measurable products formed as a consequence of O3 exposure have limited
specificity to O3, such as 4-hydroxynonenal that is formed by autoxidation, an event
that can be initiated by O3 but also by a multitude of other oxidants. Although some
classes of lipid oxidation products (e.g., specific aldehydes, cholesterol products) are
specific to O3, measurement in either BALF or in tissue does not necessarily provide
insight regarding the compartment in which they were formed (i.e., the ELF, cell
membrane, intracellular space) because the ELF is a dynamic compartment and, once
formed, hydrophobic species can partition. Oxidation of membrane components
might produce similar cellular outcomes regardless of the initiating oxidant. Lipid
ozonides, which could be generated either within the ELF or from ozonation of cell
membrane unsaturated lipids, could bind to receptors, activate signaling cascades,
and act in other ways, making differences between pure extracellular reaction and
direct membrane reaction indistinguishable. Thus, in some cases documenting
whether O3 per se reacts directly with cellular constituents might be essential
(despite the challenges of in vivo demonstrations), while in other cases precisely
where O3 reacts might be of less concern with regard to characterizing mechanisms
of health outcomes.
Thus, components of the ELF are major targets for O3 and the resulting secondary
oxidation products are key mediators of toxicity in the airways. The role of reaction
products in O3-induced toxicity is discussed in Section 5.3. The reaction cascade
resulting from the interaction of O3 with ELF substrates can then carry the oxidative
burden deeper into cells lining the RT to elicit the health effects observed.
5.2.3.1 Summary
The ELF is a complex mixture of lipids, proteins, and antioxidants that serve as the
first barrier and target for inhaled O3 (Figure 5-7). The thickness of the airways and
alveolar surface lining layers is an important determinant of the dose of O3 to the
tissues. The progressive decrease in ELF thickness and increase in interfacial surface
with progression from the TB region to the alveolus decreases the radial distance O3
or its reaction products must travel to reach the cells lining the RT. The antioxidant
substances present in the ELF appear in most cases to limit interaction of O3 with
underlying tissues and to prevent penetration of O3 deeper into the lung. However, as
the ELF thickness decreases and becomes ultra thin in the alveolar region, it may be
possible for direct interaction of O3 with the underlying epithelial cells to occur.
The formation of secondary oxidation products is likely related to the concentration
of antioxidants present and the quenching ability of the lining fluid. Mechanisms are
present to replenish the antioxidant substrate pools as well as to remove secondary
reaction products and prevent tissue interactions. Important differences exist in the
reaction rates for O3 and these ELF biomolecules and the reactivity of the resulting
products. Overall, studies suggest that UA and AH2 are more reactive with O3 than
5-27
-------
GSH, proteins, or lipids. In addition to contributing to the driving force for O3
uptake, formation of secondary oxidation products may lead to increased cellular
injury and cell signaling (discussed in Section 5.3). Studies indicate that the
antioxidants might be participating in reactions where the resulting secondary
oxidation products might penetrate into the tissue layer and lead to perturbations.
Ozone
ELF LMW Antioxidants
Uric Acid, Ascorbate,
Glutathione, a-Tocopherol
ELF Macromolecules
Surfacta nt com ponents
e.g.SP-A, phospholipids and
cholesterol,
Mucins, CCSP, Albumin,
Hyaluronan, Free fatty acids
Cellular Macromolecules
Plasma membrane proteins
and phospholipids
Free fatty acids and
carbohydrates
Mechanisms for Antioxidant
Repletion
• Secretion by epithelial cells
• Transport from plasma
• Reduction of oxidized ascorbate
Secondary Oxidation Products
Oxidized proteins
Aldehydes
Ozonized Cholesterol Species
Lipid Peroxides
Eicosanoids and PAF
Hyaluronan Fragments
Ozonized Radical
Reactive Oxygen Species
' Mechanisms for Reaction
Product Removal
• Quenching reactions by ELF
antioxidants and proteins
• Non-enzymatic reactions with
cellular antioxidants
• Metabolism by cellular GST/NQ01
• Receptor-mediated uptake by
V macrophages t
Cellular injury
Cellularsignaling
Note: Contents of this figure not discussed in Section 5.2 will be discussed in Section 5.3. Low molecular weight, LMW; Clara cell
secretory protein, CCSP; Surfactant Protein-A, SP-A; Platelet activating factor, PAF. Ozone will react with components of the ELF
to produce reaction products that may lead to cellular injury and cell signaling as discussed in Section 5.3.
Figure 5-7 Details of the O3 interaction with the airway ELF to form secondary
oxidation products.
5-28
-------
5.3 Possible Pathways/Modes of Action
5.3.1 Introduction
Mode of action refers to a sequence of key events and processes that result in a given
toxic effect (U.S. EPA. 2005). Elucidation of mechanisms provides a more detailed
understanding of these key events and processes (U.S. EPA. 2005). Moreover,
toxicity pathways describe the processes by which perturbation of normal biological
processes produce changes sufficient to lead to cell injury and subsequent events
such as adverse health effects (U.S. EPA. 2009f). The purpose of this section of
Chapter 5 is to describe the key events and toxicity pathways that contribute to health
effects resulting from short-term and long-term exposures to O3. The extensive
research carried out over several decades in humans and in laboratory animals has
yielded numerous studies on mechanisms by which O3 exerts its effects. This section
will discuss some of the representative studies with particular emphasis on studies
published since the 2006 O3 AQCD (U.S. EPA. 2006b) and on studies in humans
that inform biological mechanisms underlying responses to O3.
It is well-appreciated that secondary oxidation products, which are formed as a result
of O3 exposure, initiate numerous responses at the cellular, tissue and whole organ
level of the respiratory system. These responses include the activation of neural
reflexes, initiation of inflammation, alteration of epithelial barrier function,
sensitization of bronchial smooth muscle, modification of innate/adaptive immunity
and airways remodeling, as will be discussed below. These have the potential to
result in effects on other organ systems such as the cardiovascular, central nervous,
hepatic and reproductive systems or result in developmental effects. It has been
proposed that lipid ozonides and other secondary oxidation products, which are
bioactive and cytotoxic in the respiratory system, are responsible for systemic
effects. However it is not known whether they gain access to the vascular space
(Chuang et al.. 2009). Recent studies in animal models show that inhalation of O3
results in systemic oxidative stress. The following subsections describe the current
understanding of potential pathways and modes of action responsible for the
pulmonary and extrapulmonary effects of O3 exposure.
5.3.2 Activation of Neural Reflexes
Acute O3 exposure results in reversible effects on lung function parameters through
activation of neural reflexes. The involvement of bronchial C-fibers, a type of
nociceptive sensory nerve, has been demonstrated in dogs exposed through an
endotracheal tube to 2-3 ppm O3 for 20-70 minutes (Coleridge et al.. 1993: Schelegle
et al.. 1993). This vagal afferent pathway was found to be responsible for
O3-mediated rapid shallow breathing and other changes in respiratory mechanics in
O3-exposed dogs (Schelegle et al.. 1993). Ozone also triggers neural reflexes that
5-29
-------
stimulate the autonomic nervous system and alter electrophysiologic responses of the
heart. For example, bradycardia, altered HRV and arrhythmia have been
demonstrated in rodents exposed for several hours to 0.1-0.6 ppm O3 (Hamade and
Tankerslev. 2009: Watkinson et al.. 2001: Aritoetal.. 1990). Another effect is
hypothermia, which in rodents occurred subsequent to the activation of neural
reflexes involving the parasympathetic nervous system (Watkinson et al.. 2001).
Vagal afferent pathways originating in the RT may also be responsible for
O3-mediated activation of nucleus tractus solitarius neurons that resulted in neuronal
activation in stress-responsive regions of the central nervous system (CNS) (rats,
0.5-2.0 ppm O3 for 1.5-120 hours) (Gackiere et al.. 2011).
Recent studies in animals provide new information regarding the effects of O3 on
reflex responses mediated by bronchopulmonary C-fibers. In ex vivo mouse lungs,
O3 exposure (30 j^M solubilized) selectively activated a subset of C-fiber receptors
that are TRPA1 ion channels (Taylor-Clark and Undem, 2010). TRPA1 ion channels
are members of the TRP family of ion channels, which are known to mediate the
responses of sensory neurons to inflammatory mediators (Caceres et al., 2009).
In addition to TRPA1 ion channels possibly playing a key role in O3-induced
decrements in pulmonary function, they may mediate allergic asthma (Caceres et al.,
2009). Activation of TRPA1 ion channels following O3 exposure is likely initiated
by secondary oxidation products such as aldehydes and prostaglandins (Taylor-Clark
and Undem, 2010) through covalent modification of cysteine and lysine residues
(Trevisani et al., 2007). Ozonation of unsaturated fatty acids in the ELF was found to
result in the generation of aldehydes (Frampton et al., 1999) such as
4-hydroxynonenal and 4-oxononenal (Taylor-Clark et al., 2008: Trevisani et al.,
2007). 4-oxononenal is a stronger electrophile than 4-hydroxynonenal and exhibits
greater potency toward the TRPA1 channels (Taylor-Clark et al.. 2008: Trevisani et
al.. 2007). In addition, PGE2 is known to sensitize TRPA1 channels (Bang et al..
2007).
In humans exercising at a moderate level, the response to O3 (500 ppb for 2 h) was
characterized by substernal discomfort, especially on deep inspiration, accompanied
by involuntary truncation of inspiration (Hazucha et al.. 1989). This latter response
led to decreased inspiratory capacity and to decreased forced vital capacity (FVC)
and forced expiratory volume in one second (FEVi), as measured by spirometry.
These changes, which occurred during O3 exposure, were accompanied by decreased
VT and increased respiratory frequency in human subjects. Spirometric changes in
FEVi and FVC were not due to changes in respiratory muscle strength (Hazucha et
al.. 1989). In addition, parasympathetic involvement in the O3-mediated decreases in
lung volume was minimal (Mudway and Kelly, 2000), since changes in FVC or
symptoms were not modified by treatment with bronchodilators such as atropine in
human subjects exposed to 400 ppb O3 for 2 hours while exercising at a heavy level
(Beckett et al.. 1985). However, the loss of vital capacity was reversible with
intravenous administration of the rapid-acting opioid agonist, sufentanyl, in human
subjects exercising at a moderate level and exposed to 420 ppb O3 for 2 hours, which
indicated the involvement of opioid receptor-containing nerve fibers and/or more
central neurons (Passannante et al.. 1998). The effects of sufentanyl may be
5-30
-------
attributed to blocking C-fiber stimulation by O3 since activation of opioid receptors
downregulated C-fiber function (Belvisi et al., 1992). Thus, nociceptive sensory
nerves, presumably bronchial C-fibers, are responsible for O3-mediated responses in
humans (Passannante et al., 1998). This vagal afferent pathway is responsible for
pain-related symptoms and inhibition of maximal inspiration in humans (Hazucha et
al.. 1989).
There is some evidence that eicosanoids (see Section 5.3.3) play a role in the neural
reflex since cyclooxygenase inhibition with indomethacin (Alexis et al., 2000;
Schelegle et al., 1987) or ibuprofen, which also blocks some lipoxygenase activity
(Hazucha et al., 1996), before exposure to O3 significantly blunted the spirometric
responses. These studies involved exposures of 1-2 hours to 350-400 ppb O3 in
human subjects exercising at light, moderate, and heavy levels. In the latter study,
ibuprofen treatment resulted in measurable decreases in BALF levels of PGE2 and
TXB2 at 1-hour postexposure (Hazucha et al., 1996). Although an earlier study
demonstrated that PGE2 stimulated bronchial C-fibers (Coleridge et al., 1993;
Coleridge et al., 1976) and suggested that PGE2 mediated O3-induced decreases in
pulmonary function, no correlation was observed between the degree of
ibuprofen-induced inhibition of BALF PGE2 levels and blunting of the spirometric
response to O3 (Hazucha et al., 1996). These results point to the involvement of a
lipoxygenase product. Further, as noted above, PGE2 may play a role in the neural
reflex by sensitizing TRPA1 channels. A recent study in human subjects exercising
at a moderate to high level and exposed for 1 hour to 350 ppb O3 also provided
evidence that arachidonic acid metabolites, as well as oxidative stress, contribute to
human responsiveness to O3 (Alfaro et al., 2007).
In addition to the spirometric changes, mild airways obstruction occurred in human
subjects exercising at a moderate level during O3 exposure (500 ppb for 2 hours)
(Hazucha et al., 1989). This pulmonary function decrement is generally measured as
specific airway resistance (sRaw) which is the product of airway resistance and
thoracic gas volume. In several studies involving human subj ects exercising at a
moderate to heavy level and exposed for 1-4 hours to 200-300 ppb O3, changes in
sRaw correlated with changes in inflammatory and injury endpoints measured
18-hours postexposure, but did not follow the same time course or change to the
same degree as spirometric changes (i.e., FEVi, FVC) measured during exposure
(Balmes et al., 1996; Arisetal., 1993; Schelegle et al., 1991). In addition, a small but
persistent increase in airway resistance associated with narrowing of small peripheral
airways (measured as changes in isovolumetric FEF25_75) was demonstrated in
O3-exposed human subjects (350 ppb for 130 minutes, moderate exercise level)
(Weinmann et al., 1995c; Weinmann et al., 1995b). A similar study (400 ppb O3 for
2 hours in human subjects exercising at a heavy level) found decreases in FEF25_75
concomitant with increases in residual volume, which is suggestive of small airways
dysfunction (Kreit et al., 1989). In separate studies, a statistically significant increase
in residual volume (500 ppb for 2 hours) (Hazucha et al., 1989) and a statistically
significant decrease in FEF25_75 (160 ppb for 7.6 hours) (Horstman et al., 1995) were
observed following O3 exposure in human subjects exercising at moderate and light
5-31
-------
levels, respectively, providing further support for an O3-induced effect on small
airways.
Mechanisms underlying this rapid increase in airway resistance following O3
exposure are incompletely understood. Pretreatment with atropine decreased baseline
sRaw and prevented O3-induced increases in sRaw in human subjects exercising at a
heavy level (400 ppb for 0.5 hours) (Beckett et al., 1985), indicating the involvement
of muscarinic cholinergic receptors of the parasympathetic nervous system.
Interestingly, atropine pretreatment partially blocked the decrease in FEVi, but had
no effect on the decrease in FVC, breathing rate, tidal volume or respiratory
symptoms (Beckett et al., 1985). Using a |3-adrenergic agonist, it was shown that
smooth muscle contraction, not increased airway mucus secretion, was responsible
for O3-induced increases in airway resistance (Beckett et al., 1985). Thus, pulmonary
function decrements measured as FEVi may reflect both restrictive (such as
decreased inspiratory capacity) and obstructive (such as bronchoconstriction) type
changes in airway responses. This is consistent with findings of McDonnell et al.
(1983) who observed a relatively strong correlation between sRaw and FEVi
(r = -0.31, p = 0.001) and a far weaker correlation between sRaw and FVC (r = -0.16,
p = 0.10) in human subjects exercising at a heavy level and exposed for 2.5 hours to
120-400 ppb O3.
Furthermore, tachykinins may contribute to O3-mediated increases in airway
resistance. In addition to stimulating CNS reflexes, bronchopulmonary C-fibers
mediate local axon responses by releasing neuropeptides such as substance P (SP),
neurokinin (NK) A and calcitonin gene-related peptide (CGRP). Tachykinins bind to
NK receptors resulting in responses such as bronchoconstriction. Recent studies in
animals demonstrated that NK-1 receptor blockade had no effect on O3-stimulated
physiologic responses such as VT and fB in rats over the 8 hour exposure to 1 ppm
O3 (Oslund et al., 2008). However, SP and NK receptors contributed to vagally-
mediated bronchoconstriction in guinea pigs 3 days after a single 4-hour exposure to
2 ppm O3 (Verhein et al., 2011). In one human study in which bronchial biopsies
were performed and studied by immunohistochemistry, SP was substantially
diminished in submucosal sensory nerves 6 hours following O3 exposure (200 ppb
for 2 hours, light exercise) (Krishna et al., 1997). A statistically significant
correlation was observed between loss of SP immunoreactivity from neurons in the
bronchial mucosa and changes in FEVi measured 1-hour postexposure (Krishna et
al., 1997). Another study found that SP was increased in lavage fluid of human
subjects immediately after O3 challenge (250 ppb for 1 hour, heavy exercise)
(Hazbun et al., 1993). These results provide evidence that the increased airway
resistance observed following O3 exposure is due to vagally-mediated responses and
possibly by local axon reflex responses through bronchopulmonary C-fiber-mediated
release of SP.
A role for antioxidant defenses in modulating neural reflexes has been proposed
given the delay in onset of O3-induced pulmonary function responses that has been
noted in numerous studies. Recently, this delay was characterized in terms of
changes in fB (Schelegle et al.. 2007). In humans exposed for 1-2 hours to
5-32
-------
120-350 ppb O3 while exercising at a high level, no change in fB was observed until
a certain cumulative inhaled dose of O3 had been reached. Subsequently, the
magnitude of the change in fB was correlated with the inhaled dose rate (Schelegle et
al., 2007). These investigators proposed that initial reactions of O3 with ELF resulted
in a time-dependent depletion of ELF antioxidants, and that activation of neural
reflexes occurred only after the antioxidant defenses were overwhelmed (Schelegle et
al.. 2007).
5.3.3 Initiation of inflammation
As described previously (Section 5.2.3), O3 mainly reacts with components of the
ELF and cellular membranes resulting in the generation of secondary oxidation
products. Higher concentrations of these products may directly injure RT epithelium.
Subsequent airways remodeling may also occur (Section 5.3.7) (Mudway and Kelly,
2000). Lower concentrations of secondary oxidation products may initiate cellular
responses including cytokine generation, adhesion molecule expression, and
modification of tight junctions leading to inflammation and increased permeability
across airway epithelium (Section 5.3.4) (Dahl et al., 2007; Mudway and Kelly,
2000).
An important hallmark of acute O3 exposure in humans and animals is neutrophilic
airways inflammation. Neutrophil influx into nasal airways has been demonstrated in
human subjects (400 ppb O3 2 hours, heavy exercise) (Graham and Koren. 1990) and
in rats (0.8 ppm O3, 6 hours) (Hotchkiss et al.. 1989). Many studies of neutrophil
influx have focused on the lower airways (Hazucha et al.. 1996: Aris et al.. 1993).
The time course of this response in the lower airways and its resolution appears to be
slower than that of the decrements in pulmonary function in exercising human
subjects (Hazucha et al., 1996). In several studies, airways neutrophilia was observed
by 1-3 hours, peaked by 6 hours and was returning to baseline levels at 18-24 hours
in human subjects exercising at a heavy level and exposed for 1-2 hours to
300-400 ppb O3 (Schelegle et al.. 1991: Koren et al.. 1989: Seltzer et al.. 1986).
Neutrophils are thought to be injurious and a study in guinea pigs demonstrated that
the influx and persistence of neutrophils in airways following O3 exposure correlated
with the temporal profile of epithelial injury (0.26-1 ppm O3, 72 hours) (Hu et al.,
1982). However, neutrophils have also been shown to contribute to repair of
O3-injured epithelium in rats exposed for 8 hours to 1 ppm O3, possibly by removing
necrotic epithelial cells (Mudway and Kelly, 2000: Vesely et al., 1999). Nonetheless,
the degree of airways inflammation due to O3 is thought to have more important
long-term consequences than the more quickly resolving changes in pulmonary
function since airways inflammation is often accompanied by tissue injury (Balmes
etal.. 1996).
Ozone exposure results in alterations in other airways inflammatory cells besides
neutrophils, including lymphocytes, macrophages, monocytes and mast cells. Influx
of some of these cells accounts for the later (i.e., 18-20 hours) phase of inflammation
5-33
-------
following O3 exposure. Numbers of lymphocytes and total cells in BALF were
decreased early after O3 exposure in human subjects exercising at a light to moderate
level and exposed for 2 hours to 200 ppb O3, which preceded the neutrophil influx
(Mudwav and Kelly. 2000: Blomberg et al.. 1999: Krishna et al.. 1997). The decrease
in total cells was thought to reflect decreases in macrophages, although it was not
clear whether the cells were necrotic or whether membrane adhesive properties were
altered making them more difficult to obtain by lavage (Mudwav and Kelly. 2000:
Blomberg et al.. 1999: Mudwav et al.. 1999b: Frampton et al.. 1997b: Pearson and
Bhalla. 1997). A recent study in human subjects exercising at a moderate level and
exposed for 6.6 hours to 80 ppb O3 demonstrated an increase in numbers of sputum
monocytes and dendritic-like cells with increased expression of innate immune
surface proteins and antigen presentation markers (Alexis et al.. 2010). An increase
in submucosal mast cells was observed 1.5 hours after a 2 hour-exposure to 200 ppb
O3 (Blomberg et al.. 1999) and an increase in BAL mast cell number was observed
18 hours after a 4-hour exposure to 220 ppb O3 exposure in human subjects
exercising at a moderate level (Frampton et al.. 1997b). Mast cells may play an
important role in mediating neutrophil influx since they are an important source of
several pro-inflammatory cytokines and since their influx preceded that of
neutrophils in human subjects exercising at a moderate level and exposed for 2 hours
to 200 ppb O3 (Stenfors et al.. 2002: Blomberg et al.. 1999). Further, a study using
mast cell-deficient mice demonstrated decreased neutrophilic inflammation in
response to O3 (1.75 ppm, 3 hours) compared with wild type mice (Kleeberger et al..
1993). Influx of these inflammatory cell types in the lung is indicative of
O3-mediated activation of innate immunity as will be discussed in Section 5.3.6.
Much is known about the cellular and molecular signals involved in inflammatory
responses to O3 exposure (U.S. EPA. 2006b). Eicosanoids are one class of secondary
oxidation products that may be formed rapidly following O3 exposure and that may
mediate inflammation. Eicosanoids are metabolites of arachidonic acid—a 20-carbon
PUFA—that are released from membrane phospholipids by phospholipase
A2-mediated catalysis. Activation of phospholipase A2 occurs by several cell
signaling pathways and may be triggered by O3-mediated lipid peroxidation of
cellular membranes (Rashba-Step et al.. 1997). Additionally, cellular phospholipases
A2, C and D may be activated by lipid ozonation products (Kafoury et al.. 1998).
While the conversion of arachidonic acid to prostaglandins, leukotrienes and other
eicosanoid products is generally catalyzed by cyclooxygenases and lipoxygenases,
non-enzymatic reactions also occur during oxidative stress leading to the generation
of a wide variety of eicosanoids and reactive oxygen species. Further, the release of
arachidonic acid from phospholipids is accompanied by the formation of
lysophospholipids that are precursors for platelet activating factors (PAFs). Thus,
formation of eicosanoids, reactive oxygen species and PAFs accompanies
O3-mediated lipid peroxidation.
In addition, secondary reaction products may stimulate macrophages to produce
cytokines such as IL-1, IL-6, and TNF-a that in turn activate IL-8 production by
epithelial cells. Although IL-8 has been proposed to play a role in neutrophil
chemotaxis, measurements of IL-8 in BALF from humans exposed to O3 found
5-34
-------
increases that were too late to account for this effect (Mudway and Kelly, 2000).
The time-course profiles of PGE2 and IL-6 responses suggest that they may play a
role in neutrophil chemotaxis in humans (Mudway and Kelly, 2000). However,
pretreatment with ibuprofen attenuated O3-induced increases in BALF PGE2 levels,
but had no effect on neutrophilia in human subjects exercising at a heavy level and
exposed for 2 hour to 400 ppb O3 (Hazucha et al.. 1996).
One set of studies in humans focused on the earliest phase of airways inflammation
(1-2 hours following exposure). Human subjects, exercising at a moderate level, were
exposed to 200 ppb O3 for 2 hours and bronchial biopsy tissues were obtained 1.5
and 6 hours after exposure (Bosson et al.. 2009; Bosson et al., 2003; Stenfors et al.,
2002; Blomberg et al., 1999). Results demonstrated upregulation of vascular
endothelial adhesion molecules P-selectin and ICAM-1 at both 1.5 and 6 hours
(Stenfors et al., 2002; Blomberg et al., 1999). Submucosal mast cell numbers were
increased at 1.5 hours in the biopsy samples without an accompanying increase in
neutrophil number (Blomberg et al., 1999). Pronounced neutrophil infiltration was
observed at 6 hours in the bronchial mucosa (Stenfors et al., 2002). Surprisingly,
suppression of the NF-KB and AP-1 pathways at 1.5 hours and a lack of increased
IL-8 at 1.5 or 6 hours in bronchial epithelium were observed (Bosson et al., 2009).
The authors suggested that vascular endothelial adhesion molecules, rather than
redox sensitive transcription factors, play key roles in early neutrophil recruitment in
response to O3.
Increases in markers of inflammation occurred to a comparable degree in human
subjects with mild (least sensitive) and more remarkable (more sensitive) spirometric
responses to O3 (200 ppb, 4 hours, moderate exercise) (Balmes et al.. 1996). Two
other studies (200 ppb for 4 hours with moderate exercise and 300 ppb for 1 hour
with heavy exercise) found that acute spirometric changes were not positively
correlated with cellular and biochemical indicators of inflammation (Aris et al.,
1993; Schelegle et al., 1991). However inflammation was correlated with changes in
sRaw (Balmes et al., 1996). In another study, pretreatment with ibuprofen had no
effect on neutrophilia although it blunted the spirometric response in human subjects
exercising at heavy level and exposed for 2 hours to 400 ppb O3 (Hazucha et al.,
1996). Taken together, results from these studies indicate different mechanisms
underlying the spirometric and inflammatory responses to O3.
A common mechanism underlying both inflammation and impaired pulmonary
function was suggested by Krishna et al. (1997). This study, conducted in human
subjects exercising at a light level and exposed to 200 ppb O3 for 2 hours,
demonstrated a correlation between loss of SP immunoreactivity from neurons in the
bronchial mucosa and numbers of neutrophils and epithelial cells (shed epithelial
cells are an index of injury) in the BALF 6-hours postexposure. Furthermore, the loss
of SP immunoreactivity was correlated with the observed changes in FEVi. Another
study found that SP was increased in lavage fluid of exercising human subjects
immediately after O3 challenge (250 ppb, 1 hour, heavy exercise) (Hazbun et al..
1993). SP is a neuropeptide released by sensory nerves which mediates neurogenic
edema and bronchoconstriction (Krishna et al.. 1997). Collectively, these findings
5-35
-------
suggest that O3-mediated stimulation of sensory nerves that leads to activation of
central and local axon reflexes is a common effector pathway leading to impaired
pulmonary function and inflammation.
Studies in animal models have confirmed many of these findings and provided
evidence for additional mechanisms involved in O3-induced inflammation. A study
in mice (2 ppm O3, 3 hours) demonstrated that PAF may be important in neutrophil
chemotaxis (Longphre et al., 1999), while ICAM-1 and macrophage inflammatory
protein-2 (MIP-2), the rodent IL-8 homologue, have been implicated in a rat model
(1 ppm O3, 3 hours) (Bhalla and Gupta, 2000). Another study found that TNF
receptor, NF-KB and JNK1 mediated lung inflammation induced by O3 in mice (0.3
ppm O3, 6 and 24 hours) (Cho et al., 2007). Key roles for CXCR2, a receptor for
keratinocyte-derived chemokine (KC) and MIP-2, and for IL-6 in O3-mediated
neutrophil influx were demonstrated in mice (1 ppm O3, 3 hours) (Johnston et al.,
2005a; Johnston et al., 2005b). Activation of JNK and p38 pathways and cathepsin-S
were also found to be important in this response (3 ppm O3, 3 hours) (Williams et al.,
2009a; Williams et al., 2008a; Williams et al., 2007a). Matrix metalloproteinase-9
(MMP-9) appeared to confer protection against O3-induced airways inflammation
and injury in mice (0.3 ppm O3, 6-72 hours) (Yoon et al., 2007). Interleukin-10
(IL-10) also appeared to be protective since IL-10 deficient mice responded to O3
exposure (0.3 ppm, 24-72 hours) with enhanced numbers of BAL neutrophils,
enhanced NF-KB activation and MIP-2 levels compared with IL-10 sufficient mice
(Backus etal., 2010).
In addition, lung epithelial cells may release ATP in response to O3 exposure
(Ahmad et al., 2005). ATP and its metabolites (catalyzed by ecto-enzymes) can bind
to cellular purinergic receptors resulting in activation of cell signaling pathways
(Picher et al., 2004). One such metabolite, adenine, is capable of undergoing
oxidation leading to the formation of UA which, if present in high concentrations,
could activate inflammasomes and result in caspase 1 activation and the maturation
and secretion of IL-1(3 and IL-18 (Dostert et al., 2008). A recent study in human
subjects exercising at a moderate level and exposed for 2 hours to 400 ppb O3
demonstrated a correlation between ATP metabolites and inflammatory markers
(Esther et al., 2011), which provides some support for this mechanism.
Several recent studies have focused on the role of Toll-like receptor (TLR) and its
related adaptor protein MyD88 in mediating O3-induced neutrophilia. Hollingsworth
et al. (2004) demonstrated airways neutrophilia that was TLR4-independent
following acute (2 ppm, 3 hours) and subchronic (0.3 ppm, 72 hours) O3 exposure in
a mouse model. However, Williams et al. (2007b) found that MyD88 was important
in mediating O3-induced neutrophilia in mice (3 ppm, 3 hours), with TLR4 and
TLR2 contributing to the speed of the response. Moreover, MyD88, TLR2 and TLR4
contributed to inflammatory gene expression in this model and O3 upregulated
MyD88, TLR4 and TLR4 gene expression (Williams et al., 2007a). Neutrophilic
inflammation was also found to be partially dependent on MyD88 in mice exposed to
1 ppm O3 for 3 hours (Li et al., 2011).
5-36
-------
Hyaluronan was found to mediate a later phase (24 hours) of O3-induced
inflammation in mice (Garantziotis et al., 2010; Garantziotis et al., 2009).
Hyaluronan is an extracellular matrix component that is normally found in the ELF
as a large polymer. Exposure to 2 ppm O3 for 3 hours resulted in elevated levels of
soluble low molecular weight hyaluronan in the BALF 24-hours postexposure
(Garantziotis et al.. 2010: Garantziotis et al.. 2009). Similar results were found in
response to 3 hour exposure to 1 ppm O3 (Li et al.. 2011). Ozone may have caused
the depolymerization of hyaluronan to soluble fragments that are known to be
endogenous ligands of the CD44 receptor and TLR4 in the macrophage (Jiang et al..
2005). Binding of hyaluronan fragments to the CD44 receptor activates hyaluronan
clearance, while binding to TLR4 results in signaling through MyD88 to produce
chemokines that stimulate the influx of inflammatory cells (Jiang et al.. 2005).
Activation of NF-KB occurred in both airway epithelia and alveolar macrophages
24-hours postexposure to O3. Increases in BALF pro-inflammatory factors KC,
IL-1(3, MCP-1, TNF-a and IL-6 observed 24 hours following O3 exposure were
found to be partially dependent on TLR4 (Garantziotis et al.. 2010) while increases
in BAL inflammatory cells, which consisted mainly of macrophages, were dependent
on CD44 (Garantziotis et al.. 2009). BAL inflammatory cells number and injury
markers following O3 exposure were similar in wild-type and TLR4-deficient
animals (Garantziotis et al.. 2010).
Since exposure to O3 leads to airways inflammation characterized by neutrophilia,
and since neutrophil-derived oxidants often consume ELF antioxidants,
concentrations of ELF antioxidants have been examined during airways neutrophilia
(Long etal.. 2001: Gunnison and Hatch. 1999: Mudwav et al.. 1999b). In human
subjects exercising at a moderate level and exposed to 200 ppb O3 for 2 hours, UA,
GSH and a-TOH levels remained unchanged in BALF 6-hours postexposure while
AH2 was decreased significantly in both BALF and plasma (Mudwav et al.. 1999b).
A second study involving the same protocol reported a loss of AH2 from bronchial
wash fluid and BALF, representing proximal and distal airway ELF respectively, as
well as an increase in oxidized GSH in both compartments (Mudwav et al.. 2001).
No change was observed in ELF UA levels in response to O3 (Mudwav et al.. 2001).
Further, O3 exposure (0.8 ppm, 4 hours) in female rats resulted in a 50% decrease in
BALF AH2 immediately postexposure (Gunnison and Hatch. 1999). These studies
suggested a role for AH2 and GSH in protecting against the oxidative stress
associated with inflammation.
The relationship between inflammation, antioxidant status and O3 dose has also been
investigated. The degree of inflammation in rats has been correlated with 18O-labeled
O3 dose markers in the lower lung. In female rats exposed to 0.8 ppm O3 for 4 hours,
BAL neutrophil number and 18O reaction product were directly proportional
(Gunnison and Hatch, 1999). Kari et al. (1997) observed that a 3-week caloric
restriction (75%) in rats abrogated the toxicity of O3 (2 ppm, 2 hours), measured as
BALF increases in protein, fibronectin and neutrophils, that was seen in normally fed
rats. Accompanying this resistance to O3 toxicity was a reduction (30%) in the
accumulation of 18O reaction product in the lungs. These investigations also
demonstrated an inverse relationship between AH2 levels and O3 dose and provided
5-37
-------
evidence for AH2 playing a protective role following O3 exposure in these studies.
Pregnant and lactating rats had lower AH2 content in BALF and exhibited a greater
increase in accumulation of 18O reaction products compared with pre-pregnant rats in
response to O3 exposure (Gunni son and Hatch, 1999). In the calorie restricted model,
a 30% higher basal BALF AH2 concentration and a rapid accumulation of AH2 into
the lungs to levels 60% above normal occurred as result of O3 exposure (Kari et al..
1997). However, this relationship between AH2 levels and O3 dose did not hold up in
every study. Aging rats (9 and 24 months old) had 49% and 64% lower AH2 in lung
tissue compared with month-old rats but the aging-induced AH2 loss did not increase
the accumulation of 18O reaction products following O3 exposure (0.4-0.8 ppm,
2-6 hours) (Vincent et al.. 1996b).
A few studies have examined the dose- or concentration-responsiveness of airways
neutrophilia in O3-exposed humans (Holz et al., 1999; Devlin et al., 1991).
No concentration-responsiveness was observed in healthy human subjects exposed
for 1 hour to 125-250 ppb O3 while exercising at a light level and a statistically
significant increase in sputum neutrophilia was observed only at the higher
concentration (Holz et al., 1999). However, concentration-dependent and statistically
significant increases in BAL neutrophils and the inflammatory mediator IL-6 were
reported following exposure to 80 and 100 ppb O3 for 6.6 hours in human subjects
exercising at a moderate level (Devlin et al., 1991). Additional evidence is provided
by a meta-analysis of the O3 dose-inflammatory response in controlled human
exposure studies involving exposure to 80-600 ppb O3 for 60-396 minutes and
exercise levels ranging from light to heavy (Mudway and Kelly, 2004b). Results
demonstrated a linear relationship between inhaled O3 dose (determined as the
product of concentration, ventilation and time) and BAL neutrophils at 0-6 hours and
18-24 hours following O3 exposure (Mudwav and Kelly. 2004b).
5.3.4 Alteration of Epithelial Barrier Function
Following O3 exposure, injury and inflammation can lead to altered airway barrier
function. Histologic analysis has demonstrated damage to tight junctions between
epithelial cells, suggesting an increase in epithelial permeability. In addition, the
presence of shed epithelial cells in the BALF and increased epithelial permeability,
which is measured as the flux of small solutes, have been observed and are indicative
of epithelial injury. This could potentially lead to the loss of ELF solutes that could
diffuse down their concentration gradient from the lung to the blood. Increases in
vascular permeability, as measured by BALF protein and albumin, have also been
demonstrated (Costa etal., 1985; Huetal, 1982).
An early study in sheep measured changes in airway permeability as the flux of
inhaled radiolabeled histamine into the plasma (Abraham et al.. 1984). Exposure of
sheep to 0.5 ppm O3 for 2 hours via an endotracheal tube resulted in an increased rate
of histamine appearance in the plasma at 1 day postexposure. Subsequently,
numerous studies have measured epithelial permeability as the flux of the small
5-38
-------
solute 99mTc-DTPA that was introduced into the air spaces in different regions of the
RT. Increased pulmonary epithelial permeability, measured as the clearance of
99mTc-DTPA from lung to blood, was demonstrated in humans 1-2 hours following a
2-hour exposure to 400 ppb O3 while exercising at a heavy level (Kehrl et al, 1987).
Another study in human subjects found increased epithelial permeability 19-hours
postexposure to 240 ppb O3 for 130 minutes while exercising at moderate level
(Foster and Stetkiewicz. 1996). Increased bronchial permeability was also observed
in dogs 1-day postexposure (0.4 ppm O3 by endotracheal tube for 6 hours) and did
not resolve for several days (Foster and Freed. 1999).
A role for tachykinins in mediating airway epithelial injury and decreased barrier
function has been suggested. Nishiyama et al. (1998) demonstrated that capsaicin,
which depletes nerve fibers of substance P, blocked the O3-induced increase in
permeability of guinea pig tracheal mucosa (0.5-3 ppm O3, 0.5 hours). Pretreatment
with propranolol or atropine failed to inhibit this response, suggesting that adrenergic
and cholinergic pathways were not involved. In another study, tachykinins working
through NK-1 and CGRP receptors were found to contribute to airway epithelial
injury in O3-exposed rats (1 ppm, 8 hours) (Oslund et al., 2009, 2008).
Kleeberger et al. (2000) evaluated genetic susceptibility to O3-induced altered barrier
function in recombinant inbred strains of mice. Lung hyperpermeability, measured as
BALF protein, was evaluated 72 hours after exposure to 0.3 ppm O3 and found to be
associated with a functioning Tlr4 gene. This study concluded that Tlr4 was a strong
candidate gene for susceptibility to hyperpermeability in response to O3 (Kleeberger
et al.. 2000). A subsequent study by these same investigators found that Tlr4
modulated mRNA levels of the Nos2 genes and suggested that the protein product of
Nos2, iNOS, plays an important role in O2-induced lung hyperpermeability (0.3 ppm,
72 hours) (Kleeberger et al., 2001). More recently, HSP70 was identified as part of
the TLR4 signaling pathway (0.3 ppm, 6-72 hours) (Bauer et al., 2011).
Antioxidants have been shown to confer resistance to O3-induced injury. In a recent
study, lung hyperpermeability in response to O3 (0.3 ppm, 48 hours) was
unexpectedly reduced in mice deficient in the glutamate-cysteine ligase modifier
subunit gene compared with sufficient mice (Johansson et al.. 2010). Since the lungs
of these mice exhibited 70% glutathione depletion, protection against O3-induced
injury was unexpected (Johansson et al.. 2010). However it was found that several
other antioxidant defenses, including metallothionein, were upregulated in response
to O3 to a greater degree in the glutathione-deficient mice compared with sufficient
mice (Johansson et al.. 2010). The authors suggested that resistance to O3-induced
lung injury was due to compensatory augmentation of antioxidant defenses
(Johansson et al.. 2010). Antioxidant effects have also been attributed to Clara cell
secretory protein (CCSP) and surfactant protein A (SP-A). CCSP was found to
modulate the susceptibility of airway epithelium to injury in mice exposed to O3 (0.2
or 1 ppm for 8 hours) by an unknown mechanism (Plopper et al.. 2006). SP-A
appeared to confer protection against O3-induced airways inflammation and injury in
mice (2 ppm, 3 hours) (Hague et al.. 2007).
5-39
-------
Increased epithelial permeability has been proposed to play a role in allergic
sensitization (Matsumura, 1970), in activation of neural reflexes and in stimulation of
smooth muscle receptors (Dimeo et al., 1981). Abraham et al. (1984) reported a
correlation between airway permeability and airways hyperresponsiveness (AHR) in
Os-exposed sheep. However a recent study in human subjects exposed to 220 ppb O3
for 135 minutes while exercising at a light to moderate level did not find a
relationship between O3-induced changes in airway permeability and AHR (Que et
5.3.5 Sensitization of Bronchial Smooth Muscle
Bronchial reactivity is generally determined in terms of a response to a challenge
agent. Non-specific bronchial reactivity in humans is assessed by measuring the
effect of inhaling increasing concentrations of a bronchoconstrictive drug on lung
mechanics (sRaw or FEVi). Methacholine is most commonly employed but
histamine and other agents are also used. Specific bronchial reactivity is assessed by
measuring effects in response to an inhaled allergen in individuals (or animals)
already sensitized to that allergen. An increase in sRaw in response to non-specific or
specific challenge agents indicates AHR.
In addition to causing mild airways obstruction as discussed above, acute O3
exposure results in reversible increases in bronchial reactivity by mechanisms that
are not well understood. In one study, bronchial reactivity of healthy subjects was
significantly increased 19-hours postexposure to O3 (120-240 ppb O3 for 2 hours
with moderate exercise) (Foster et al.. 2000). These effects may be more
considerable in human subjects with already compromised airways (Section 5.4.2.2).
Ozone may sensitize bronchial smooth muscle to stimulation through an exposure-
related effect on smooth muscle or through effects on the sensory nerves in the
epithelium or on the motor nerves innervating the smooth muscle (O'Byrne et al..
1984: O'Byrne et al.. 1983: Holtzman et al.. 1979). It is also recognized that
increased bronchial reactivity can be both a rapidly occurring and a persistent
response to O3 (Foster and Freed. 1999). Tachykinins and secondary oxidation
products of O3 have been proposed as mediators of the early response and
inflammation-derived products have been proposed as mediators of the later response
(Foster and Freed. 1999). Furthermore, bronchial reactivity may be increased as a
result of O3-mediated generation of ROS.
Ozone-induced increases in epithelial permeability, which could improve access of
agonist to smooth muscle receptors, may be one mechanism of sensitization through
a direct effect on bronchial smooth muscle (Holtzman et al., 1979). As noted above, a
correlation between airway permeability and AHR has been reported in O3-exposed
sheep (Abraham et al., 1984) but not in O3-exposed human subjects (Que et al.,
2011).
5-40
-------
Neurally-mediated sensitization has been demonstrated. In human subjects exposed
for 2 hours to 600 ppb O3 while exercising at a light level, pretreatment with atropine
inhibited O3-induced AHR, suggesting the involvement of cholinergic postganglionic
pathways (Holtzman et al., 1979). Animal studies have demonstrated that O3-induced
AHR involved vagally-mediated responses (rabbits, 0.2 ppm O3, 72 hours) (Freed et
al.. 1996) and local axon reflex responses through bronchopulmonary C-fiber-
mediated release of SP (guinea pigs, 0.8 ppm O3, 2 hours) (Joad et al.. 1996).
Further, pretreatment with capsaicin to deplete nerve fibers of SP blocked
O3-mediated AHR (guinea pigs, 1-2 ppm O3, 2-2.25 hours) (Tepper et al.. 1993).
Other investigators demonstrated that SP released from airway nociceptive neurons
in ferrets contributed to O3-induced AHR (2 ppm O3, 3 hours) (Wu et al.. 2008c: Wu
et al.. 2003).
Some evidence suggests the involvement of arachidonic acid metabolites and
neutrophils in mediating O3-induced AHR (Seltzer et al., 1986; Fabbri et al., 1985).
Increased BAL neutrophils and cyclooxygenase products were found in one study
demonstrating AHR in human subjects exercising at a heavy level immediately
postexposure to 600 ppb O3 for 2 hours (Seltzer et al., 1986). Another study found
that ibuprofen pretreatment had no effect on AHR in human subjects exercising at a
heavy level following exposure to 400 ppb O3 for 2 hours, although spirometric
responses were blunted (Hazucha et al., 1996). This study measured arachidonic acid
metabolites and provided evidence that that the arachidonic acid metabolites whose
generation was blocked by ibuprofen, (i.e., prostaglandins, thromboxanes and some
leukotrienes) did not play a role in AHR. Experiments in dogs exposed for 2 hours to
2.1 ppm O3 demonstrated a close correlation between O3-induced AHR and airways
neutrophilic inflammation measured in tissue biopsies (Holtzman et al., 1983).
Furthermore, the increased AHR observed in dogs following O3 exposure (3 ppm,
2 hours) was inhibited by neutrophil depletion (O'Byrne et al.. 1983) and by pre-
treatment with inhibitors of arachidonic acid metabolism. In one of these studies,
indomethacin pre-treatment did not prevent airways neutrophilia in response to O3
(3 ppm, 2 hours) providing evidence that the subset of arachidonic acid metabolites
whose generation was inhibitable by the cyclooxygenase inhibitor indomethacin
(i.e., prostaglandins and thromboxanes) was not responsible for neutrophil influx
(O'Byrne et al.. 1984). It should be noted that these studies did not measure whether
the degree to which the inhibitor blocked arachidonic acid metabolism and thus their
results should be interpreted with caution. Taken together, these findings suggest that
arachidonic acid metabolites may be involved in the AHR response following O3
exposure in dogs. Studies probing the role of neutrophils in mediating the AHR
response have provided inconsistent results (Al-Hegelan et al.. 2011).
Evidence for cytokine and chemokine involvement in the AHR response to O3 has
been described. Some studies have suggested a role for TNF-a (mice, 0.5 and 2 ppm
O3, 3 hours) (Cho et al., 2001; Shore et al., 2001) and IL-1 (mice and ferrets, 2 ppm
O3, 3 hours) (Wu et al.. 2008c; Park et al.. 2004). The latter study found that SP
expression in airway neurons was upregulated by IL-1 that was released in response
to O3. Other studies in mice have demonstrated a key role for CXCR2, the
chemokine receptor for the neutrophil chemokines KC and MIP-2, but not for IL-6 in
5-41
-------
O3-mediated AHR (1 ppm O3, 3 hours) (Johnston et al., 2005a; Johnston et al.,
2005b). In contrast, CXCR2 and IL-6 were both required for neutrophil influx in this
model (Johnston et al., 2005a; Johnston et al., 2005b), as discussed above. Williams
et al. (2008b) demonstrated that the Th2 cytokine IL-13 contributed to AHR, as well
as to airways neutrophilia, in mice (3 ppm O3, 3 hours).
Other studies have focused on the role of TLR4. Hollingsworth et al. (2004)
measured AHR, as well as airways neutrophilia, in mice 6 and 24 hours following
acute (2 ppm O3 for 3 hours) and subchronic (0.3 ppm for 3 days) exposure to O3.
TLR4 is a key component of the innate immune system and is responsible for the
immediate inflammatory response seen following challenge with endotoxin and other
pathogen-associated substances. In this study, a functioning TLR4 was required for
the full AHR response following O3 exposure but not for airways neutrophilia
(Hollingsworth et al., 2004). These findings are complemented by an earlier study
demonstrating that O3 effects on lung hyperpermeability required a functioning
TLR4 (mice, 0.3 ppm O3, 72 hours) (Kleeberger et al.. 2000). Williams et al. (2007b)
found that TLR2, TLR4 and the TLR adaptor protein MyD88 contributed to AHR in
mice (3 ppm O3, 3 hours). Ozone was also found to upregulate MyD88, TLR4 and
TLR4 gene expression in this model (Williams et al., 2007b). Furthermore, a recent
study reported O3-induced AHR that required TLR4 and MyD88 in mice exposed to
1 ppm O3 for 3 hours (Li et al., 2011).
A newly recognized mechanistic basis for O3-induced AHR is provided by studies
focusing on the role of hyaluronan following O3 exposure in mice (Garantziotis et
al.. 2010: Garantziotis et al.. 2009). Hyaluronan is an extracellular matrix component
that is normally found in the ELF as a large polymer. Briefly, TLR4 and CD44 were
found to mediate AHR in response to O3 and hyaluronan. Exposure to 2 ppm O3 for
3 hours resulted in enhanced AHR and elevated levels of soluble low molecular
weight hyaluronan in the BALF 24-hours postexposure (Garantziotis et al., 2010;
Garantziotis et al., 2009). Ozone may have caused the depolymerization of
hyaluronan to soluble fragments that are known to be endogenous ligands of the
CD44 receptor and TLR4 in the macrophage (Jiang et al., 2005). In the two recent
studies, O3-induced AHR was attenuated in CD44 and TLR4-deficient mice
(Garantziotis et al., 2010; Garantziotis et al., 2009). Hyaluronan fragment-mediated
stimulation of AHR was found to require functioning CD44 receptor and TLR4
(Garantziotis et al., 2010; Garantziotis et al., 2009). In contrast, high-molecular-
weight hyaluronan blocked AHR in response to O3 (Garantziotis et al., 2009).
In another study high-molecular-weight hyaluronan enhanced repair of epithelial
injury (Jiang et al., 2005). These studies provide a link between innate immunity and
the development of AHR following O3 exposure, and indicate a role for TLR4 in
increasing airways responsiveness. While TLR4-dependent responses usually involve
activation of NF-KB and the upregulation of proinflammatory factors, the precise
mechanisms leading to AHR are unknown (Al-Hegelan et al., 2011).
In guinea pigs, AHR was found to be mediated by different pathways at 1- and
3-days postexposure to a single exposure of O3 (2 ppm for 4 hours) (Verhein et al..
2011; Yost et al.. 2005). At 1 day, AHR was due to activation of airway
5-42
-------
parasympathetic nerves rather than to an exposure-related effect on smooth muscle
(Yost et al., 2005). This effect occurred as a result of O3-stimulated release of major
basic protein from eosinophils (Yost et al., 2005). Major basic protein is known to
block inhibitory M2 muscarinic receptors that normally dampen acetylcholine release
from parasympathetic nerves (Yost et al., 2005). The resulting increase in
acetylcholine release caused an increase in smooth muscle contraction following O3
exposure (Yost et al.. 2005). Eosinophils played a different role 3-days postexposure
to O3 in guinea pigs (Yost et al.. 2005). Ozone-mediated influx of eosinophils into
lung airways resulted in a different population of cells present 3-days postexposure
compared to those present at 1 day (Yost et al.. 2005). At this time point, eosinophil-
derived major basic protein increased smooth muscle responsiveness to acetylcholine
which also contributed to AHR (Yost et al.. 2005). However, the major effect of
eosinophils was to protect against vagal hyperreactivity (Yost et al.. 2005).
The authors suggested that these beneficial effects were due to the production of
nerve growth factor (Yost et al.. 2005). Further work by these investigators
demonstrated a key role for IL-1(3 in mediating AHR 3-days postexposure to O3
(Verhein et al.. 2011). In this study, IL-1(3 increased nerve growth factor and SP that
acted through the NK1 receptor to cause vagally-mediated bronchoconstriction
(Verhein et al.. 2011). The mechanism by which SP caused acetylcholine release
from parasympathetic nerves following O3 exposure was not determined (Verhein et
al.. 2011). Taken together, the above study results indicate that mechanisms involved
in O3-mediated AHR can vary over time postexposure and that eosinophils and SP
can play a role. Results of this animal model may provide some insight into allergic
airways disease in humans that is characterized by eosinophilia (Section 5.4.2.2).
5.3.6 Modification of Innate/Adaptive Immune System Responses
Host defense depends on effective barrier function and on innate immunity and
adaptive immunity (Al-Hegelan et al.. 2011). The effects of O3 on barrier function in
the airways were discussed above (Section 5.3.4). This section focuses on the
mechanisms by which O3 impacts innate and adaptive immunity. Both tissue damage
and foreign pathogens are triggers for the activation of the innate immune system.
This results in the influx of inflammatory cells such as neutrophils, mast cells,
basophils, eosinophils, monocytes and dendritic cells and the generation of cytokines
such as TNF-a, IL-1, IL-6, KC and IL-17. Further, innate immunity encompasses the
actions of complement and collections, and the phagocytic functions of macrophages,
neutrophils and dendritic cells. Airway epithelium also contributes to innate immune
responses. Innate immunity is highly dependent on cell signaling networks involving
TLR4. Adaptive immunity provides immunologic memory through the actions of B
and T-cells. Important links between the two systems are provided by dendritic cells
and antigen presentation. Recent studies demonstrate that exposure to O3 modifies
cells and processes which are required for innate immunity, contributes to innate-
adaptive immune system interaction and primes pulmonary immune responses to
endotoxin.
5-43
-------
Ozone exposure of human subjects resulted in recruitment of activated innate
immune cells to the airways. Healthy individuals were exposed to 80 ppb O3 for
6.6 hours while exercising at a moderate level and airways inflammation was
characterized in induced sputum 18-hours postexposure (Alexis et al.. 2010).
Previous studies demonstrated that induced sputum contains liquid and cellular
constituents of the ELF from central conducting airways (Alexis et al.. 2001b) and
also identified these airways as a site of preferential O3 absorption during exercise
(Hu et al.. 1994). Ozone exposure resulted in increased numbers of neutrophils,
airway monocytes and dendritic-like cells in sputum (Alexis et al.. 2010). In addition,
increased expression of cell surface markers characteristic of innate immunity and
antigen presentation (i.e., CD-14 and HLA-DR) was demonstrated on airway
monocytes (Alexis et al.. 2010). Enhanced antigen presentation contributes to
exaggerated T-cell responses and promotes Th2 inflammation and an allergic
phenotype (Lav et al.. 2007). Upregulation of pro-inflammatory cytokines was also
demonstrated in sputum of O3-exposed subjects (Alexis et al.. 2010). One of these
cytokines, IL-12p70, correlated with numbers of dendritic-like cells in the sputum,
and is an indicator of dendritic cell activation (Alexis et al.. 2010). These authors
have previously reported that exposure of human subjects exercising at a light to
moderate level to 400 ppb O3 for 2 hours resulted in activation of monocytes and
macrophages (Lay et al.. 2007). which could play a role in exacerbating existing
asthma by activating allergen-specific memory T-cells. The current study confirms
these findings and extends them by suggesting a potential mechanism whereby
O3-activated dendritic cells could stimulate naive T-cells to promote the
development of asthma (Alexis et al.. 2010). A companion study by these same
investigators (described in detail in Section 5.4.2.1) provides evidence of dendritic
cell activation, measured as increased expression of HLA-DR, in a subset of the
human subjects (GSTM1 null) exposed to 400 ppb O3 for 2 hours while exercising at
a light to moderate level (Alexis et al., 2009). Since dendritic cells are a link between
innate and adaptive immunity, these studies provide evidence for an O3-mediated
interaction between the innate and adaptive immune systems.
Another recent study linked O3-mediated activation of the innate immune system to
the development of non-specific AHR in a mouse model (Pichavant et al.. 2008).
Repeated exposure to 1 ppm O3 for 3 hours (3 days over a 5 day period) induced
non-specific AHR measured 24 hours following the last exposure (Pichavant et al..
2008). This response was found to require NKT-cells, which are effector
lymphocytes of innate immunity, as well as IL-17 and airways neutrophilia
(Pichavant et al.. 2008). Since glycolipids such as galactosyl ceramide are ligands for
the invariant GDI receptor on NKT-cells and serve as endogenous activators of
NKT-cells, a role for O3-oxidized lipids in activating NKT-cells was proposed
(Pichavant et al.. 2008). The authors contrasted this innate immunity pathway with
that of allergen-provoked specific AHR which involves adaptive immunity, the
cytokines IL-4, IL -13, IL-17, and airways eosinophilia (Pichavant et al.. 2008).
Interestingly, NKT-cells were required for both the specific AHR provoked by
allergen and the non-specific AHR provoked by O3 (Pichavant et al.. 2008).
Different cytokine profiles of the NKT-cells from allergen and O3-exposed mice
were proposed to account for the different pathways (Pichavant et al.. 2008). More
5-44
-------
recently, NKT-cells have been found to function in both innate and adaptive
immunity (Vivier et al., 2011).
An interaction between allergen and O3 in the induction of nonspecific AHR was
shown in another animal study (Larsen et al.. 2010). Mice were sensitized with the
aerosolized allergen OVA on 10 consecutive days followed by exposure to O3
(0.1-0.5 ppm for 3 hours) (Larsen et al., 2010). While allergen sensitization alone did
not alter airways responsiveness to a nonspecific challenge, O3 exposure of
sensitized mice resulted in nonspecific AHR at 6- and 24-hours postexposure (Larsen
et al., 2010). The effects of O3 on AHR were independent of airways eosinophilia
and neutrophilia (Larsen et al., 2010). However, OVA pretreatment led to goblet cell
metaplasia which was enhanced by O3 exposure (Larsen et al., 2010). It should be
noted that OVA sensitization using only aerosolized antigen in this study is less
common than the usual procedure for OVA sensitization achieved by one or more
initial systemic injections of OVA and adjuvant followed by repeated inhalation
exposure to OVA. This study also points to an interaction between innate and
adaptive immune systems in the development of the AHR response.
Furthermore, O3 was found to act as an adjuvant for allergic sensitization
(Hollingsworth et al.. 2010). Oropharyngeal aspiration of OVA on day 0 and day 6
failed to lead to allergic sensitization unless mice were first exposed to 1 ppm O3 for
2 hours (Hollingsworth et al.. 2010). The O3-mediated response involved Th2 (IL-4,
IL-5 and IL-9) and ThlV cytokines (IL-17) and was dependent on a functioning
TLR4 (Hollingsworth et al.. 2010). Ozone exposure also activated OVA-bearing
dendritic cells in the thoracic lymph nodes, as measured by the presence of the CD86
surface marker, which suggests naive T-cell stimulation and the involvement of Th2
pathways (Hollingsworth et al.. 2010). Thus the adjuvant effects of O3 may be due to
activation of both innate and adaptive immunity.
Priming of the innate immune system by O3 was reported by Hollingsworth et al.
(2007). In this study, exposure of mice to 2 ppm O3 for 3 hours led to nonspecific
AHR at 24- and 48-hours postexposure, an effect which subsided by 72 hours
(Hollingsworth et al.. 2007). However, in mice treated with aerosolized endotoxin
immediately following O3 exposure, AHR was greatly enhanced at 48-and 72-hours
postexposure (Hollingsworth et al.. 2007). In addition, O3 pre-exposure was found to
reduce the number of inflammatory cells in the BALF, to increase cytokine
production and total protein in the BALF and to increase systemic IL-6 following
exposure to endotoxin (Hollingsworth et al.. 2007). Furthermore, O3 stimulated the
apoptosis of alveolar macrophages 24-hours postexposure, an effect which was
greatly enhanced by endotoxin treatment. Apoptosis of circulating blood monocytes
was also observed in response to the combined exposures (Hollingsworth et al..
2007). Ozone pre-exposure enhanced the response of lung macrophages to endotoxin
(Hollingsworth et al.. 2007). Taken together, these findings demonstrated that O3
exposure increased innate immune responsiveness to endotoxin. The authors
attributed these effects to the increased surface expression of TLR4 and increased
signaling in macrophages observed in the study (Hollingsworth et al.. 2007). It was
proposed that the resulting decrease in airway inflammatory cells could account for
5-45
-------
O3-mediated decreased clearance of bacterial pathogens observed in numerous
animal models (Hollingsworth et al., 2007).
More recently, these authors demonstrated that hyaluronan contributed to the
O3-primed response to endotoxin (Li et al.. 2010). In this study, exposure of mice to
1 ppm O3 for 3 hours resulted in enhanced responses to endotoxin, which was
mimicked by intratracheal instillation of hyaluronan fragments (Li et al., 2010).
Hyaluronan, like O3, was also found to induce TLR4 receptor peripheralization in the
macrophage membrane (Li et al., 2010; Hollingsworth et al., 2007), an effect which
is associated with enhanced responses to endotoxin. This study and previous ones by
the same investigators showed elevation of BALF hyaluronan in response to O3
exposure (Garantziotis et al., 2010; Li et al., 2010; Garantziotis et al., 2009),
providing evidence that the effects of O3 on innate immunity are at least in part
mediated by hyaluronan fragments. The authors note that excessive TLR4 signaling
can lead to lung injury and suggest that O3 may be responsible for an exaggerated
innate immune response which may underlie lung injury and decreased host defense
(LietaL2010).
Activation or upregulation of the immune system has not been reported in all studies.
Impaired antigen-specific immunity was demonstrated following subacute O3
exposure (0.6 ppm, 10 h/day for 15 days) in mice (Feng et al.. 2006). Specifically, O3
exposure altered the lymphocyte subset and cytokine profile and impacted thymocyte
early development leading to immune dysfunction. Further, recent studies
demonstrated SP-A oxidation in mice exposed for 3-6 hours to 2 ppm O3. SP-A is an
important innate immune protein which plays a number of roles in host defense
including acting as opsonin for the recognition of some pathogens (Hague et al..
2009). These investigations found that O3-mediated carbonylation of purified SP-A
was associated with impaired macrophage phagocytosis in vitro (Mikerov et al.,
2008c). In addition, O3 exposure (2 ppm for 3 hours) in mice was found to increase
susceptibility to pneumonia infection in mice through an impairment of SP-A
dependent phagocytosis (Mikerov et al., 2008b; Mikerov et al., 2008a). Furthermore,
early life exposure to O3 in infant monkeys followed by a recovery period led to
hyporesponsiveness to endotoxin (Maniar-Hew et al., 2011), as discussed below and
in Section 5.4.2.4 and Section 7.2.3.2.
Taken together, results of recent studies provide evidence that O3 alters host
immunologic response and leads to immune system dysfunction through its effects
on innate and adaptive immunity.
5.3.7 Airways Remodeling
The nasal airways, conducting airways and distal airways (i.e., respiratory
bronchioles or CAR depending on the species) have all been identified as sites of
O3-mediated injury and inflammation (Mudway and Kelly, 2000). At all levels of the
RT, loss of sensitive epithelial cells, degranulation of secretory cells, proliferation of
resistant epithelial cells and neutrophilic influx have been observed as a result of O3
5-46
-------
exposure (Mudway and Kelly, 2000; Cho et al., 1999). An important study (Plopper
et al., 1998) conducted in adult rhesus monkeys (0.4 and 1.0 ppm O3 for 2 hours at
rest) found that 1 ppm O3 resulted in the greatest epithelial injury in the respiratory
bronchioles immediately postexposure although injury was observed at all of the RT
sites studied except for the lung parenchyma. Exposure to 0.4 ppm O3 resulted in
epithelial injury only in the respiratory bronchioles. Initial cellular injury correlated
with site-specific O3 dose since the respiratory bronchioles received the greatest O3
dose (18O mass/lung weight) and sustained the greatest initial cellular injury.
The respiratory bronchioles were also the site of statistically significant GSH
reduction. In addition, a study in isolated perfused rat lungs found greater injury in
conducting airways downstream of bifurcations where local doses of O3 were higher
(Postlethwait et al.. 2000).
In addition to the degree of initial injury, the degree of airways inflammation due to
O3 may have important long-term consequences since airways inflammation may
lead to tissue injury (Balmes et al., 1996). Persistent inflammation and injury,
observed in animal models of chronic and intermittent exposure to O3, are associated
with airways remodeling, including mucous cell metaplasia of nasal transitional
epithelium (Harkema et al., 1999; Hotchkiss et al., 1991) and bronchiolar metaplasia
of alveolar ducts (Mudway and Kelly, 2000). In a nonhuman primate model,
hyperplasia of both URT and LRT epithelium resulted from chronic exposure to O3
concentrations as low as 0.15 ppm and 0.3 ppm (Harkema et al., 1993, 1987b).
Fibrotic changes such as deposition of collagen in the airways and sustained lung
function decrements especially in small airways have also been demonstrated as a
response to chronic O3 exposure (Mudway and Kelly, 2000; Chang et al., 1992).
These effects, described in detail in Section 7.2.3.2, have been demonstrated in rats
exposed to levels of O3 as low as 0.25 ppm. Mechanisms responsible for the
resolution of inflammation and the repair of injury remain to be clarified and there is
only a limited understanding of the biological processes underlying long-term
morphological changes. However, a recent study in mice demonstrated a key role for
the TGF-(3 signaling pathway in the deposition of collagen in the airways wall
following chronic intermittent exposure to 0.5 ppm O3 (Katre et al.. 2011). Studies in
infant monkeys have also documented effects of chronic intermittent exposure to
0.5 ppm O3 on the developing lung and immune system. Extensive discussion of this
topic is found in Section 5.4.2.4 (Lifestage) and in Section 7.2.3.2.
It should be noted that repeated exposure to O3 results in attenuation of some
O3-induced responses, including those associated with the activation of neural
reflexes (e.g., decrements in pulmonary function), as discussed in Section 5.3.2.
However, numerous studies demonstrate that some markers of injury and
inflammation remain increased during multi-day exposures to O3. Mechanisms
responsible for attenuation, or the lack thereof, are incompletely understood.
5-47
-------
5.3.8 Systemic Inflammation and Oxidative/Nitrosative Stress
Extrapulmonary effects of O3 have been noted for decades (U.S. EPA. 2006b). It has
been proposed that lipid oxidation products resulting from reaction of O3 with lipids
in the ELF are responsible for systemic effects, however it is not known whether they
gain access to the vascular space (Chuang et al., 2009). Alternatively,
extrapulmonary release of diffusible mediators may initiate or propagate
inflammatory responses in the vascular or systemic compartments (Cole and
Freeman, 2009). A role for O3 in modulating endothelin, a potent vasoconstrictor,
has also been proposed. Studies in rats found that exposure to 0.4 and 0.8 ppm O3
induced endothelin system genes in the lung and increased circulating levels of
endothelin (Thomson et al., 2006; Thomson et al., 2005). Systemic oxidative stress
may be suggested by studies in humans which reported associations between O3
exposure and levels of plasma 8-isoprostanes and the presence of peripheral blood
lymphocyte micronuclei (Chen et al., 2007a; Chen et al., 2006a). However, plasma
isoprostanes are not a direct measure of systemic oxi dative stress since they are
stable and can be generated in any compartment before diffusion into the vascular
space. Evidence of O3-mediated systemic oxidative stress is better provided by new
animal studies described below.
Ozone-induced perturbations of the cardiovascular system were recently investigated
in young mice and monkeys (Chuang et al.. 2009) and in rats (Kodavanti et al.. 2011:
Perepu et al.. 2010) (see Section 6.3.3 and Section 7.3.1.2). These are the first studies
to suggest that systemic oxidative stress and inflammation play a mechanistic role in
O3-induced effects on the systemic vasculature and heart. Exposure to 0.5 ppm O3
for 5 days resulted in oxidative/nitrosative stress, vascular dysfunction and
mitochondrial DNA damage in the aorta (Chuang et al., 2009). Chronic exposure to
0.8 ppm O3 resulted in an enhancement of inflammation and lipid peroxidation in the
heart following an ischemia-reperfusion challenge (Perepu et al., 2010). In addition,
chronic intermittent exposure to 0.4 ppm O3 increased aortic levels of mRNA for
biomarkers of oxidative stress, thrombosis, vasoconstriction and proteolysis and
aortic lectin-like oxidized-low density lipoprotein receptor-1 (LOX-1) mRNA and
protein levels (Kodavanti et al., 2011). The latter study suggests a role for circulating
oxidized lipids in mediating the effects of O3.
Two recent controlled human exposure studies also demonstrated cardiovascular
effects in response to short-term O3 exposure (see Section 6.3.1). Changes in high
frequency heart rate variability (HRV) were reported, albeit effects were in opposing
directions (Devlin et al.. 2012: Fakhri et al.. 2009). Differences in study design may
account for this discrepancy; the increase in high frequency HRV was observed
following relatively low O3 exposure (120 ppb for 2 hours during rest) and the
decrease in high frequency HRV was observed following higher O3 exposures (300
ppb for 2 hours while exercising at a high rate). These changes in cardiac function
provide evidence of O3-induced modulation of the autonomic nervous system,
potentially through the activation of neural reflexes in the lung. Devlin et al. (2012)
also demonstrated O3-induced increases in markers of systemic inflammation and a
pro-thrombogenic environment (Devlin et al., 2012). Older controlled human
5-48
-------
exposure studies have reported increased myocardial work (Gong et al., 1998), and
increased markers of oxidative stress (Liu et al., 1999; Liu et al., 1997) following a
single O3 exposure and reduced serum tocopherol levels following repeated O3
exposures (Foster et al., 1996) (see Section 6.3.1). These findings in humans,
together with findings from animal toxicological studies, provide evidence of O3-
mediated cardiovascular effects that may involve changes in autonomic tone,
systemic and/or vascular oxidative stress and inflammation, and activation of the
fibrinolytic system.
Systemic inflammation and oxidative/nitrosative stress may similarly affect other
organ systems as well as the plasma compartment. Circulating cytokines have the
potential to enter the brain through diffusion and active transport and to contribute to
neuroinflammation, neurotoxicity, cerebrovascular damage and a break-down of the
blood brain barrier (Block and Calderon-Garciduenas, 2009) (see Section 6.4 and
Section 7.5). They can also activate neuronal afferents (Block and Calderon-
Garciduefias, 2009). Vagal afferent pathways originating in the RT may also be
responsible for O3-mediated activation of nucleus tractus solitarius neurons which
resulted in neuronal activation in stress-responsive regions of the CNS in rats (0.5 or
2 ppm O3 for 1.5-120 hours) (Gackiere et al., 2011). Recent studies have
demonstrated O3-induced brain lipid peroxidation, cytokine production in the brain
and upregulated expression of VEGF in rats (0.5 ppm O3, 3 hours or 0.25-0.5 ppm
O3, 4 h/day, 15-60 days) (Guevara-Guzman et al.. 2009: Aranedaet al.. 2008:
Pereyra-Mufioz et al., 2006). Further, O3-induced oxidative stress resulted in
increased plasma lipid peroxides (0.25 ppm, 4h/day, 15-60 days) (Santiago-Lopez et
al., 2010), which was correlated with damage to specific brain regions (Pereyra-
Mufioz et al., 2006).
Oxidative stress is one mechanism by which testicular and sperm function may be
disrupted (see Section 7.4.1). Studies in Leydig cells in vitro have demonstrated that
steroidogenesis is blocked by oxidative stress (Diemer et al., 2003). It has been
proposed that lipid peroxidation of sperm plasma membrane may lead to impaired
sperm mobility and decreased sperm quality (Agarwal et al., 2003). Further, it has
been proposed that oxidative stress may damage DNA in the sperm nucleus and lead
to apoptosis and a decline in sperm counts (Agarwal et al., 2003). One study reported
an association between O3 exposure and semen quality and suggested oxidative
stress as an underlying mechanism (Sokol et al., 2006). Additional evidence is
required to substantiate this link.
A role for plasma antioxidants in modulating O3-induced respiratory effects was
suggested by a recent study (Aibo et al., 2010). In this study, pretreatment of rats
with a high dose of acetaminophen resulted in increased levels of plasma cytokines
and the influx of inflammatory cells into the lung following O3 exposure
(0.25-0.5 ppm, 6 hours) (Aibo et al., 2010). These effects were not observed in
response to O3 alone. Furthermore, acetaminophen-induced liver injury was
exacerbated by O3 exposure. A greater increase in hepatic neutrophil accumulation
and greater alteration in gene expression profiles was observed in mice exposed to
O3 and acetaminophen compared with either exposure alone (Aibo et al., 2010).
5-49
-------
Although not measured in this study, glutathione depletion in the liver is known to
occur in acetaminophen toxicity. Since liver glutathione is the source of plasma
glutathione, acetaminophen treatment may have lowered plasma glutathione levels
and altered the redox balance in the vascular compartment. These findings indicate
interdependence between RT, plasma and liver responses to O3, possibly related to
glutathione status.
5.3.9 Impaired Alveolar-Arterial Oxygen Transfer
Ozone may impair alveolar-arterial oxygen transfer and reduce the supply of arterial
oxygen to the myocardium. This may have a greater impact in individuals with
compromised cardiopulmonary systems. Gong et al. (1998) provided evidence of a
small decrease in arterial oxygen saturation in human subjects exposed for 3 hours to
300 ppb O3 while exercising at a light to moderate level. In addition, Delaunois et al.
(1998) demonstrated pulmonary vasoconstriction in O3-exposed rabbits (0.4 ppm,
4 hours). Although of interest, the contribution of this pathway to O3-induced
cardiovascular effects remains uncertain.
5.3.10 Summary
This section summarizes the modes of action and toxicity pathways resulting from
O3 inhalation (Figure 5-8). These pathways provide a mechanistic basis for the health
effects which are described in detail in Chapters 6 and 7. However the precise
sequence by which the key events lead to health effects is not entirely clear. Three
distinct short-term responses have been well-characterized in humans challenged
with O3: decreased pulmonary function, airways inflammation, and increased
bronchial reactivity. In addition, O3 exposure exacerbates, and possibly also causes,
asthma and allergic airways disease in humans. Animal studies have demonstrated
airways remodeling and fibrotic changes in response to chronic and intermittent O3
exposures and a wide range of other responses. While the RT is the primary target
tissue, cardiovascular and other organ effects occur following short- and long-term
exposures of animals to O3.
The initial key event in the toxicity pathway of O3 is the formation of secondary
oxidation products in the RT. This mainly involves direct reactions with components
of the ELF. The resulting secondary oxidation products transmit signals to the
epithelium, nociceptive sensory nerve fibers and, if present, dendritic cells, mast cells
and eosinophils. Thus, the effects of O3 are mediated by components of ELF and by
the multiple cell types found in the RT. Further, oxidative stress is an implicit part of
this initial key event.
Another key event in the toxicity pathway of O3 is the activation of neural reflexes
which lead to decrements in pulmonary function (see Section 6.2.1). Evidence is
accumulating that secondary oxidation products are responsible for this effect.
5-50
-------
Eicosanoids have been implicated in humans, while both eicosanoids and aldehydes
are effective in animal models. Different receptors on bronchial C-fibers have been
shown to mediate separate effects of O3 on pulmonary function. Nociceptive sensory
nerves are involved in the involuntary truncation of inspiration which results in
decreases in FVC, FEVi, tidal volume and pain upon deep inspiration. Opioids block
these responses while atropine has only a minimal effect. New evidence in an animal
model suggests that TRPA1 receptors on bronchial C-fibers mediate this pathway.
Ozone exposure also results in activation of vagal sensory nerves and a mild increase
in airway obstruction measured as increased sRaw. Atropine and (3-adrenergic
agonists greatly inhibit this response in humans indicating that the airways
obstruction is due to bronchoconstriction. Other studies in humans implicated SP
release from bronchial C-fibers resulting in airway narrowing due to either
neurogenic edema or bronchoconstriction. New evidence in an animal model
suggests that the SP-NK receptor pathway caused bronchoconstriction following O3
exposure. Activation of neural reflexes also results in extrapulmonary effects such as
bradycardia.
Mode of Action/Possible Pathways
Ozone + Respiratory Tract
i
Formation of secondary oxidation products
I
Activation
of neural
reflexes
Initiation of
inflammation
Sensitization
of bronchial
smooth muscle
Systemic inflammation and
oxidative/nitrosative stress
Extrapulmonary Effects
Decrements in pulmonary function
Pulmonary inflammation/oxidative stress
Increases in airways permeability
Airways hyperresponsiveness
Exacerbation/induction of asthma
Decreased host defenses
Epithelial metaplasia and fibrotic changes
Altered lung development
Figure 5-8 The modes of action/possible pathways underlying the health
effects resulting from inhalation exposure to Os.
5-51
-------
Initiation of inflammation is also a key event in the toxicity pathway of O3.
Secondary oxidation products, as well as chemokines and cytokines elaborated by
airway epithelial cells and macrophages, have been implicated in the initiation of
inflammation. Vascular endothelial adhesion molecules may also play a role. Work
from several laboratories using human subjects and animal models suggest that O3
triggers the release of tachykinins such as SP from airway sensory nerves which
could contribute to downstream effects including inflammation (see Section 6.2.3
and Section 7.2.4). Airways neutrophilia has been demonstrated in BALF, mucosal
biopsy and induced sputum samples. Influx of mast cells, monocytes and
macrophages also occur. Inflammation further contributes to O3-mediated oxidative
stress. Recent investigations show that O3 exposure leads to the generation of
hyaluronan fragments from high molecular weight polymers of hyaluronan normally
found in the ELF in mice. Hyaluronan activates TLR4 and CD44-dependent
signaling pathways in macrophages, and results in an increased number of
macrophages in the BALF. Activation of these pathways occurs later than the acute
neutrophilic response suggesting that they may contribute to longer-term effects of
O3. The mechanisms involved in clearing O3-provoked inflammation remain to be
clarified. It should be noted that inflammation, as measured by airways neutrophilia,
is not correlated with decrements in pulmonary function as measured by spirometry.
A fourth key event in the toxicity pathway of O3 is alteration of epithelial barrier
function. Increased permeability occurs as a result of damage to tight junctions
between epithelial cells subsequent to O3-induced injury and inflammation. It may
play a role in allergic sensitization and in AHR (see Section 6.2.2, Section 6.2.6, and
Section 7.2.5). Tachykinins mediate this response while antioxidants may confer
protection. Genetic susceptibility has been associated with functioning Tlr4 and Nos2
genes.
A fifth key event in the toxicity pathway of O3 is the sensitization of bronchial
smooth muscle. Increased bronchial reactivity can be both a rapidly occurring and a
persistent response. The mechanisms responsible for early and later AHR are not
well-understood (see Section 6.2.2). One proposed mechanism of sensitization,
O3-induced increases in epithelial permeability, would improve access of agonist to
smooth muscle receptors. The evidence for this mechanism is not consistent. Another
proposed mechanism, for which there is greater evidence, is neurally-mediated
sensitization. In humans exposed to O3, atropine blocked the early AHR response
indicating the involvement of cholinergic postganglionic pathways. Animal studies
demonstrated that O3-induced AHR involved vagally-mediated responses and local
axon reflex responses through bronchopulmonary C-fiber-mediated release of SP.
Later phases of increased bronchial reactivity may involve the induction of IL-1(3
which in turn upregulates SP production. In guinea pigs, eosinophil-derived major
basic protein contributed to the stimulation of cholinergic postganglionic pathways.
A novel role for hyaluronan in mediating the later phase effects O3-induced AHR has
recently been demonstrated. Hyaluronan fragments stimulated AHR in a TLR4- and
CD44 receptor-dependent manner. Tachykinins and secondary oxidation products of
O3 have been proposed as mediators of the early response and inflammation-derived
products have been proposed as mediators of the later response. Inhibition of
5-52
-------
arachidonic acid metabolism was ineffective in blocking O3-induced AHR in humans
while in animal models mixed results were found. Other cytokines and chemokines
have been implicated in the AHR response to O3 in animal models.
A sixth key event in the toxicity pathway of O3 is the modification of innate/adaptive
immunity. While the majority of evidence for this key event comes from animal
studies, there are several studies suggesting that this pathway may also be relevant in
humans. Ozone exposure of human subjects resulted in recruitment of activated
innate immune cells to the airways. This included macrophages and monocytes with
increased expression of cell surface markers characteristic of innate immunity and
antigen presentation, the latter of which could contribute to exaggerated T-cell
responses and the promotion of an allergic phenotype. Evidence of dendritic cell
activation was observed in GSTM1 null human subjects exposed to O3, suggesting
O3-mediated interaction between the innate and adaptive immune systems. Animal
studies further linked O3-mediated activation of the innate immune system to the
development of nonspecific AHR, demonstrated an interaction between allergen and
O3 in the induction of nonspecific AHR, and found that O3 acted as an adjuvant for
allergic sensitization through the activation of both innate and adaptive immunity.
Priming of the innate immune system by O3 was reported in mice. This resulted in an
exaggerated response to endotoxin which included enhanced TLR4 signaling in
macrophages. Ozone-mediated impairment of the function of SP-A, an innate
immune protein, has also been demonstrated. Taken together these studies provide
evidence that O3 can alter host immunologic response and lead to immune system
dysfunction. These mechanisms may underlie the exacerbation and induction of
asthma (see Section 6.2.6 and Section 7.2.1), as well as decreases in host defense
(see Section 6.2.5 and Section 7.2.6).
Another key event in the toxicity pathway of O3 is airways remodeling. Persistent
inflammation and injury, which are observed in animal models of chronic and
intermittent exposure to O3, are associated with morphologic changes such as
mucous cell metaplasia of nasal epithelium, bronchiolar metaplasia of alveolar ducts
and fibrotic changes in small airways (see Section 7.2.3). Mechanisms responsible
for these responses are not well-understood. However a recent study in mice
demonstrated a key role for the TGF-(3 signaling pathway in the deposition of
collagen in the airway wall following chronic intermittent exposure to O3. Chronic
intermittent exposure to O3 has also been shown to result in effects on the developing
lung and immune system.
Systemic inflammation and vascular oxidative/nitrosative stress are also key events
in the toxicity pathway of O3. Extrapulmonary effects of O3 occur in numerous organ
systems, including the cardiovascular, central nervous, reproductive, and hepatic
systems (see Sections 6.3 to 6.5 and Sections 7.3 to 7.5). It has been proposed that
lipid oxidation products resulting from reaction of O3 with lipids and/or cellular
membranes in the ELF are responsible for systemic responses; however, it is not
known whether they gain access to the vascular space. Alternatively, release of
diffusible mediators from the lung into the circulation may initiate or propagate
inflammatory responses in the vascular or in systemic compartments.
5-53
-------
5.4 Interindividual Variability in Response
Responses to O3 exposure are variable within the population (Mudway and Kelly,
2000). Some studies have shown a large range of pulmonary function responses to
O3 among healthy young adults (i.e., 4 hours to 200 ppb O3 or for 1.5 hours to
420 ppb O3 while exercising at a moderate level) (Hazucha et al., 2003; Balmes et
al., 1996). Since individual responses were relatively consistent across time in these
studies, it was thought that responsiveness reflected an intrinsic characteristic of the
subject (Mudwav and Kelly. 2000). Other studies have shown that age and body
mass index may influence responsiveness to O3. In human subjects exercising
moderately and exposed to 420 ppb O3 for 1.5 hours, older adults were generally not
responsive to O3 (Hazucha et al.. 2003). while obese young women appeared to be
more responsive than lean young women (Bennett et al.. 2007). In another study, the
observed lack of spirometric responsiveness in one group of human subjects was not
attributable to the presence of endogenous endorphins, which could vary between
individuals and which could potentially block C-fiber stimulation by O3 (420 ppb,
2 hours, moderate exercise (Passannante et al.. 1998). Other responses to O3 have
also been characterized by a large degree of interindividual variability. For example,
interindividual variability in the neutrophilic response has been noted in human
subjects (Holzetal., 1999: Devlin etal, 1991: Schelegle et al.. 1991). One study
demonstrated a 3-fold difference in airways neutrophilia, measured as percent of total
cells in proximal BALF, among human subjects exposed to 300 ppb O3 for 1 hour
while exercising at a heavy level (Schelegle et al.. 1991). Another study reported a
20-fold difference in BAL neutrophils following exposure to 80-100 ppb O3 for
6.6 hours in human subjects exercising at a moderate level (Devlin et al.. 1991).
In contrast, reproducibility of intraindividual responses to 1-hour exposure to
250 ppb O3 in human subjects exercising at a light level, measured as sputum
neutrophilia, was demonstrated by Holz et al. (1999). While the basis for the
observed interindividual variability in responsiveness to O3 is not clear, both
dosimetric and mechanistic factors are likely to contribute and will be discussed
below.
5.4.1 Dosimetric Considerations
Two studies have investigated the correlation of O3 uptake with the pulmonary
function responses to O3 exposure (Reeser et al.. 2005: Gerritv et al.. 1994). These
studies found that the large subject-to-subject variability in %AFEVi response to O3
does not appear to have a dosimetric explanation. Reeser et al. (2005) found no
significant relationship between %AFEVi and fractional absorption of O3 using the
bolus method. Contrary to previous findings, the percent change in dead space
volume of the respiratory tract (%AVD) did not correlate with O3 uptake, possibly
due to the contraction of dead space caused by airway closure. Gerritv et al. (1994)
found that intersubject variability in FEVi and airway resistance was not related to
differences in the O3 dose delivered to the lower airways, whereas minute ventilation
5-54
-------
was predictive of FEVi decrement. No study has yet demonstrated that subjects show
a consistent pattern of O3 retention when re-exposed over weeks of time, as has been
shown to be the case for the FEVi response, or that within-subject variation in FEVi
response is related to fluctuations in O3 uptake. However, these studies did not
control for the differences in conducting airway volume between individuals.
By controlling for conducting airway volume, it may be possible to estimate how
much of the intersubject variation in FEVi response at a given O3 exposure is due to
actual inter-individual variability in dose.
5.4.2 Mechanistic Considerations
This section considers mechanistic factors that may contribute to variability in
responses between individuals. It has been proposed that some of the variability may
be genetically determined (Yang et al., 2005a). Besides gene-environment
interactions, other factors such as pre-existing diseases and conditions, nutritional
status, lifestage, attenuation, and co-exposures may also contribute to inter-individual
variability in responses to O3 and are discussed below.
5.4.2.1 Gene-environment Interactions
The pronounced interindividual variation in responses to O3 infers that genetic
background may play an important role in responsiveness to O3 (Cho and
Kleeberger. 2007: Kleeberger et al.. 1997) (see also Section 8.4). Strains of mice
which are prone or resistant to O3-induced effects have been used to systematically
identify candidate susceptibility genes. Using these recombinant inbred strains of
mice and exposures to 0.3 ppm O3 for up to 72 hours, genome wide linkage analyses
(also known as positional cloning) demonstrated quantitative trait loci for
O3-induced lung inflammation and hyperpermeability on chromosome 17
(Kleeberger et al., 1997) and chromosome 4 (Kleeberger et al., 2000), respectively.
More specifically, these studies found that Tnf, whose protein product is the
inflammatory cytokine TNF-a, and Tlr4, whose protein product is TLR4, were
candidate susceptibility genes (Kleeberger et al., 2000; Kleeberger et al., 1997).
Other studies, which used targeted deletion, identified genes encoding iNOS and heat
shock proteins as TLR4 effector genes (Bauer et al., 2011; Kleeberger et al., 2001)
and found that IL-10 protects against O3-induced pulmonary inflammation (Backus
et al., 2010). Investigations in inbred mouse strains found that differences in
expression of certain proteins, such as CCSP (1.8 ppm O3 for 3 hours) (Broeckaert et
al.. 2003) and MARCO (0.3 ppm O3 for up to 48 hours) (Dahl et al.. 2007). were
responsible for phenotypic characteristics, such as epithelial permeability and
scavenging of oxidized lipids, respectively, which confer sensitivity to O3.
Genetic polymorphisms have received increasing attention as modulators of
O3-mediated effects. Functionally relevant polymorphisms in candidate susceptibility
genes have been studied at the individual and population level in humans, and also in
5-55
-------
animal models. Genes whose protein products are involved in antioxidant
defense/oxidative stress and xenobiotic metabolism, such as glutathione-S-
transferase Ml (GSTM1) and NADPH: quinone oxidoreductase 1 (NQO1), have also
been a major focuses of these efforts. This is because oxidative stress resulting from
O3 exposure is thought to contribute to the pathogenesis of asthma, and because
xenobiotic metabolism detoxifies secondary oxidation products formed by O3 which
contribute to oxi dative stress (Islam et al.. 2008). TNF-a is of interest since it is
linked to a candidate O3 susceptibility gene and since it plays a key role in initiating
airways inflammation (Li et al.. 2006d). Polymorphisms of genes coding for
GSTM1, NQO1 and TNF-a have been associated with altered risk of O3-mediated
effects (Li et al.. 2006d: Yang et al.. 2005a: Romieu et al.. 2004b: Corradi et al..
2002: Bergamaschi et al.. 2001). Additional studies have focused on functional
variants in other genes involved in antioxidant defense such as catalase (CAT),
myeloperoxidase, heme oxygenase (HMOX-1) and manganese superoxide dismutase
(MnSOD) (Wenten et al.. 2009: Islam et al.. 2008). These studies are discussed
below.
GSTM1 is a phase II antioxidant enzyme which is transcriptionally regulated by
NF-e2-related factor 2-antioxidant response element (Nrf2-ARE) pathway. A large
proportion (40-50%) of the general public (across ethnic populations) has the
GSTMl-null genotype, which has been linked to an increased risk of health effects
due to exposure to air pollutants (London, 2007). A role for GSTs in metabolizing
electrophiles such as 4-hydroxynonenal, which is a secondary oxidation product
resulting from O3 exposure, has been demonstrated (Awasthi et al., 2004). A recent
study found that the GSTM1 genotype modulated the time course of the neutrophilic
inflammatory response following acute O3 exposure (400 ppb for 2 hours with light
to moderate exercise) in healthy adults (Alexis et al.. 2009). In GSTMl-null and -
sufficient subjects, O3-induced sputum neutrophilia was similar at 4 hours. However,
neutrophilia resolved by 24 hours in sufficient subjects but not in GSTMl-null
subjects. In contrast, no differences in 24 hour sputum neutrophilia were observed
between GSTMl-null and -sufficient human subjects exposed to 60 ppb O3 for 6.6
hours with moderate exercise (Kim et al.. 2011). It is not known whether the effect
seen at the higher exposure level (Alexis et al.. 2009) was due to the persistence of
pro-inflammatory stimuli, impaired production of downregulators or impaired
neutrophil apoptosis and clearance. However, a subsequent in vitro study by these
same investigators found that GSTM1 deficiency in airway epithelial cells enhanced
IL-8 production in response to 0.4 ppm O3 for 4 hours (Wu et al.. 2011).
Furthermore, NF-KB activation was required for O3-induced IL-8 production (Wu et
al.. 2011). Since IL-8 is a potent neutrophil activator and chemotaxin, this study
provides additional evidence for the role of GSTM1 as a modulator of inflammatory
responses due to O3 exposure.
In addition, O3 exposure increased the expression of the surface marker CD 14 in
airway neutrophils of GSTMl-null subjects to a greater extent than in sufficient
subjects (Alexis et al., 2009). Furthermore, differences in airway macrophages were
noted between the GSTM1-sufficient and -null subjects. Numbers of airway
macrophages were decreased at 4 and 24 hours following O3 exposure in
5-56
-------
GSTM1-sufficient subjects (Alexis et al., 2009). Airway macrophages in
GSTMl-null subjects were greater in number and found to have greater oxidative
burst and phagocytic capability than those of sufficient subjects. Airway
macrophages and dendritic cells from GSTMl-null subjects exposed to O3 expressed
higher levels of the surface marker HLA-DR, suggesting activation of the innate
immune system (Alexis et al.. 2009). These differences in inflammatory responses
between the GSTMl-null and -sufficient subjects may provide biological plausibility
for the differences in O3-mediated effects reported in controlled human exposure
studies (Corradi et al.. 2002: Bergamaschi et al.. 2001). It should also be noted that
GSTM1 genotype did not affect the acute pulmonary function (i.e., spirometric)
response to O3 which provides additional evidence for separate mechanisms
underlying the effects of O3 on pulmonary function and inflammation in adults
(Alexis et al.. 2009). However, GSTMl-null asthmatic children were previously
found to be more at risk of O3-induced effects on pulmonary function than
GSTM1-sufficient asthmatic children (Romieu et al.. 2004b).
Another enzyme involved in the metabolism of secondary oxidation products is
NQO1. NQO1 catalyzes the 2-electron reduction by NADPH of quinones to
hydroquinones. Depending on the substrate, it is capable of both protective
detoxification reactions and redox cycling reactions resulting in the generation of
reactive oxygen species. A recent study using NQOl-null mice demonstrated that
NQO1 contributes to O3-induced oxidative stress, AHR and inflammation following
a 3-hour exposure to 1 ppm O3 (Voynow et al., 2009). These experimental results
may provide biological plausibility for the increased biomarkers of oxidative stress
and increased pulmonary function decrements observed in O3-exposed individuals
bearing both the wild-type NQO1 gene and the null GSTM1 gene (Corradi et al.,
2002: Bergamaschi et al.. 2001).
Besides enzymatic metabolism, other mechanisms participate in the removal of
secondary oxidation products formed as a result of O3 inhalation. One involves
scavenging of oxidized lipids via the macrophage receptor with collagenous structure
(MARCO) expressed on the cell surface of alveolar macrophages. A recent study
demonstrated increased gene expression of MARCO in the lungs of an O3-resistant
C3H mouse strain (HeJ) but not in an O3-sensitive, genetically similar strain (OuJ)
(Dahl et al., 2007). Upregulation of MARCO occurred in mice exposed to 0.3 ppm
O3 for 24-48 hours; inhalation exposure for 6 hours at this concentration was
insufficient for this response. Animals lacking the MARCO receptor exhibited
greater inflammation and injury, as measured by BAL neutrophils, protein and
isoprostanes, following exposure to 0.3 ppm O3 (Dahl et al., 2007). MARCO also
protected against the inflammatory effects of oxidized surfactant lipids (Dahl et al.,
2007). Scavenging of oxidized lipids may limit O3-induced injury since ozonized
cholesterol species formed in the ELF (mice, 0.5-3 ppm O3, 3 hours) (Pulfer et al.,
2005; Pulfer and Murphy, 2004) stimulated apoptosis and cytotoxicity in vitro (Gao
et al., 2009b; Sathishkumar et al., 2009; Sathishkumar et al., 2007b; Sathishkumar et
al.. 2007a).
5-57
-------
Two studies reported relationships between TNF promoter variants and O3-induced
effects in humans. In one study, O3-induced change in lung function was
significantly lower in adult subjects with TNF promoter variants -308A/A and -
308G/A compared with adult subjects with the variant -308G/G (Yang et al., 2005a).
This response was modulated by a specific polymorphism of LTA (Yang et al.,
2005a), a previously identified candidate susceptibility gene whose protein product is
lymphotoxin-a (Kleeberger et al.. 1997). In the second study, an association between
the TNF promoter variant -308G/G and decreased risk of asthma and lifetime
wheezing in children was found (Li et al.. 2006d). The protective effect on wheezing
was modulated by ambient O3 levels and by GSTM1 and GSTP1 polymorphisms.
The authors suggested that the Z7VF-308 G/G genotype may have a protective role in
the development of childhood asthma (Li et al.. 2006d).
Similarly, a promoter variant of the gene HMOX-1, consisting of a smaller number of
(GT)n repeats, was associated with a reduced risk for new-onset asthma in non-
Hispanic white children (Islam et al., 2008). The number of (GT)n repeats in this
promoter has been shown to be inversely related to the inducibility of HMOX-1.
A modulatory effect of O3 was demonstrated since the beneficial effects of this
polymorphism were seen only in children living in low O3 communities (Islam et al.,
2008). This study also identified an association between a polymorphism of the CAT
gene and increased risk of new-onset asthma in Hispanic children; however no
modulation by O3 was seen (Islam et al., 2008). No association was observed in this
study between aMnSOD polymorphism and asthma (Islam et al., 2008).
Studies to date indicate that some variability in individual responsiveness to O3 may
be accounted for by functional genetic polymorphisms. Further, the effects of
gene-environment interactions may be different in children and adults.
5.4.2.2 Pre-existing Diseases and Conditions
Pre-existing diseases and conditions can alter the response to O3 exposure. For
example, responsiveness to O3, as measured by spirometry, is decreased in smokers
and individuals with COPD (U.S. EPA. 2006b). Asthma and allergic diseases are of
major importance in this discussion. In individuals with asthma, there is increased
responsiveness to bronchoconstrictor challenge. This results from a combination of
structural and physiological factors including increased airway inner-wall thickness,
smooth muscle responsiveness and mucus secretion. Although inflammation is likely
to contribute, its relationship to AHR is not clear (U.S. EPA. 2006b). However, some
asthmatics have higher baseline levels of neutrophils, lymphocytes, eosinophils and
mast cells in bronchial washes and bronchial biopsy tissue (Stenfors et al.. 2002).
It has been proposed that enhanced sensitivity to O3 is conferred by the presence of
greater numbers of resident airway inflammatory cells in disease states such as
asthma flVIudway and Kelly. 2000).
In order to determine whether asthmatics exhibit greater responses to O3, several
earlier studies compared pulmonary function in asthmatic and non-asthmatic subj ects
5-58
-------
following O3 exposure. Some also probed mechanisms which could account for
enhanced sensitivity. While the majority focused on measurements of FEVi and
FVC and found no differences between the two groups following exposures of
2-4 hours to 125-250 ppb O3 or to a 30-minute exposure to 120-180 ppb O3 by
mouthpiece in human subjects exercising at a light to moderate level (Stenfors et al..
2002: Mudwavetal.. 2001: Holzetal.. 1999: Scannell et al.. 1996: Koenig et al..
1987: Linn et al.. 1978). there were notable exceptions. In one study, greater airways
obstruction in asthmatics compared with non-asthmatic subjects was observed
immediately following a 2-hour exposure to 400 ppb O3 while exercising at a heavy
level (Kreit et al.. 1989). These changes were measured as statistically significant
greater decreases in FEVi and in FEF25-75 (but not in FVC) in the absence of a
bronchoconstrictor challenge (Kreit et al.. 1989). These results suggest that this
group of asthmatics responded to O3-exposure with a greater degree of vagally-
mediated bronchoconstriction compared with the non-asthmatics. A second study
demonstrated a statistically significant greater decrease in FEVi and in FEVi/FVC
(but not in FVC) in asthmatics compared with non-asthmatics exposed to 160 ppb O3
for 7.6 hours with light exercise (Horstman et al.. 1995). These responses were
accompanied by wheezing and inhaler use in the asthmatics (Horstman et al.. 1995).
Aerosol bolus dispersion measurements demonstrated a statistically significant
greater change in asthmatics compared with non-asthmatics, which was suggestive of
O3-induced small airway dysfunction (Horstman et al.. 1995). Furthermore, a
statistically significant correlation was observed between the degree of baseline
airway status and the FEVi response to O3 in the asthmatic subjects (Horstman et al..
1995). A third study found similar decreases in FVC and FEVi in both asthmatics
and non-asthmatics exposed to 400 ppb O3 for 2 hours with light exercise (Alexis et
al.. 2000). However, a statistically significant decrease in FEF75, a measure of small
airway function, was observed in asthmatics but not in non-asthmatics (Alexis et al..
2000). Taken together, these latter studies indicate that while the magnitude of
restrictive type spirometric decline was similar in asthmatics and non-asthmatics, that
obstructive type changes (i.e., bronchoconstriction) were greater in asthmatics.
Further, asthmatics exhibited greater sensitivity to O3 in terms of small airways
function.
Since asthma exacerbations occur in response to allergens and/or other triggers, some
studies have focused on O3-induced changes in AHR following a bronchoconstrictor
challenge. No difference in sensitivity to methacholine bronchoprovocation was
observed between asthmatics and non-asthmatics exposed to 400 ppb O3 for 2 hours
while exercising at a heavy level (Kreit et al.. 1989). However, increased bronchial
reactivity to inhaled allergens was demonstrated in mild allergic asthmatics exposed
to 160 ppb for 7.6 hours, 250 ppb for 3 hours and 120 ppb for 1 hour while
exercising at a light level or at rest (Kehrl et al.. 1999: Jorres et al.. 1996: Molfino et
al.. 1991) and in allergen-sensitized guinea pigs following O3 exposure (1 ppm,
1 hour) (Sun et al.. 1997). Similar, but modest, responses were reported for
individuals with allergic rhinitis (Jorres et al.. 1996). Further, the contractile response
of isolated airways from human donor lung tissue, which were sensitized and
challenged with allergen, was increased by pre-exposure to 1 ppm O3 for 20 (Roux et
5-59
-------
al., 1999). These studies provide support for O3-mediated enhancement of responses
to allergens in allergic subjects.
In terms of airways neutrophilia, larger responses were observed in asthmatics
compared to non-asthmatics subjects, who were exercising at a light to moderate
level and exposed to O3, in some (Balmes et al.. 1997: Scannell et al.. 1996: Basha et
al., 1994) but not all (Mudway et al., 2001) of the earlier studies. While each of these
studies involved exposure of exercising human subjects to 200 ppb O3, the duration
of exposure was longer (i.e., 4-6 hours) in the former studies than in the latter study
(2 hours). Further, statistically significantly increases in myeloperoxidase levels (an
indicator of neutrophil activation) in bronchial washes was observed in mild
asthmatics compared with non-asthmatics, despite no difference in O3-stimulated
neutrophil influx between the 2 groups following exposure to 200 ppb O3 for 2 hours
with moderate exercise (Stenfors et al., 2002). A more recent study found that atopic
asthmatic subjects exhibited an enhanced inflammatory response to O3 (400 ppb,
4 hours, with light to moderate exercise) (Hernandez et al., 2010). This response was
characterized by greater numbers of neutrophils, higher levels of IL-6, IL-8 and
IL-1(3 and greater macrophage cell-surface expression of TLR4 and IgE receptors in
induced sputum compared with healthy subjects. This study also reported a greater
increase in hyaluronan in atopic subjects and atopic asthmatics compared with
healthy subjects following O3 exposure. Animal studies have previously reported that
hyaluronic acid activates TLR4 signaling and results in AHR (see Section 5.3.5).
Furthermore, levels of IL-10, a potent anti-inflammatory cytokine, were greatly
reduced in atopic asthmatics compared to healthy subjects. These results provide
evidence that innate immune and adaptive responses are different in asthmatics and
healthy subjects exposed to O3.
Eosinophils may be an important modulator of responses to O3 in asthma and allergic
airways disease. Eosinophils and associated proteins are thought to affect muscarinic
cholinergic receptors which are involved in vagally-mediated bronchoconstriction
(Mudway and Kelly, 2000). Studies described in Section 5.3.5 which demonstrated a
key role of eosinophils in O3-mediated AHR may be relevant to human allergic
airways disease which is characterized by airways eosinophilia (Yost et al., 2005).
Furthermore, O3 exposure sometimes results in airways eosinophilia in allergic
subjects or animal models. For example, eosinophilia of the nasal and other airways
was observed in individuals with pre-existing allergic disease following O3
inhalation (160 ppb, 7.6 hours with light exercise and 270 ppb, 2 hours with
moderate exercise) (Vagaggini et al., 2002: Peden et al., 1997). Further, O3 exposure
(0.5 ppm, 8 hours/day for 1-3 days) increased allergic responses, such as eosinophilia
and augmented intraepithelial mucosubstances, in the nasal airways of ovalbumin
(OVA)-sensitized rats (Wagner et al., 2002). In contrast, Stenfors et al. (2002) found
no stimulation of eosinophil influx measured in bronchial washes and BALF of mild
asthmatics following exposure to a lower concentration (200 ppb, 2 hours, with
moderate exercise) of O3.
The role of mast cells in O3-mediated asthma exacerbations has been investigated.
Mast cells are thought to play a key role in O3-induced airways inflammation, since
5-60
-------
airways neutrophilia was decreased in mast cell-deficient mice exposed to O3
(Kleeberger et al., 1993). However, another study found that mast cells were not
involved in the development of increased bronchial reactivity in O3-exposed mice
(Noviski et al., 1999). Nonetheless, mast cells release a wide variety of important
inflammatory mediators which may lead to asthma exacerbations (Stenfors et al.,
2002). A large increase in mast cell number in bronchial submucosa was observed in
non-asthmatics and a significant decrease in mast cell number in bronchial
epithelium was observed in mild asthmatics 6 hours following exposure to 200 ppb
O3 for 2 hours during mild exercise (Stenfors et al.. 2002). While these results point
to an O3-mediated flux in bronchial mast cell populations which differed between the
non-asthmatics and mild asthmatics, interpretation of these findings is difficult.
Furthermore, mast cell number did not change in airway lavages in either group in
response to O3 (Stenfors et al.. 2002)
Cytokine profiles in the airways have been investigated as an indicator of O3
sensitivity. Differences in epithelial cytokine expression were observed in bronchial
biopsy samples in non-asthmatic and asthmatic subjects both at baseline and 6-hours
postexposure to 200 ppb O3 for 2 hours with moderate exercise (Bosson et al., 2003).
The asthmatic subjects had a higher baseline expression of IL-4 and IL-5 compared
to non-asthmatics. In addition, expression of IL-5, IL-8, GM-CSF, and ENA-78 in
asthmatics was increased significantly following O3 exposure compared to non-
asthmatics (Bosson et al., 2003). Some of these (IL-4, IL-5 and GM-CSF) are
Th2-related cytokines or neutrophil chemoattractants, and play a role in IgE
production, airways eosinophilia and suppression of Thl-cytokine production
(Bosson et al., 2003). These findings suggest a link between adaptive immunity and
enhanced responses of asthmatics to O3.
A further consideration is the compromised status of ELF antioxidants in disease
states such as asthma (Mudway and Kelly, 2000). This could possibly be due to
ongoing inflammation which causes antioxidant depletion or to abnormal antioxidant
transport or synthesis (Mudway and Kelly, 2000). For example, basal levels of AH2
were significantly lower and basal levels of oxidized GSH and UA were significantly
higher in bronchial wash fluid and BALF of mild asthmatics compared with healthy
control subjects (Mudway et al., 2001). Differences in ELF antioxidant content have
also been noted between species. These observations have led to the suggestion that
the amount and composition of ELF antioxidants, the capacity to replenish
antioxidants in the ELF or the balance between beneficial and injurious interactions
between antioxidants and O3 may contribute to O3 sensitivity, which varies between
individuals and species (Mudway et al., 2006; Mudway and Kelly, 2000; Mudway et
al., 1999a). The complexity of these interactions was demonstrated by a study in
which a 2-hour exposure to 200 ppb O3, while exercising at a moderate level,
resulted in similar increases in airway neutrophils and decreases in pulmonary
function in both mild asthmatics and healthy controls, despite differences in ELF
antioxidant concentrations prior to O3 exposure (Mudway et al., 2001). Further, the
O3-induced increase in oxidized GSH and decrease in AH2 observed in ELF of
healthy controls was not observed in mild asthmatics (Mudway et al., 2001). While
the authors concluded that basal AH2 and oxidized GSH concentrations were not
5-61
-------
predictive of responsiveness to O3, they also suggested that the increased basal UA
concentrations in the mild asthmatics may have played a protective role (Mudway et
al., 2001). Thus compensatory mechanisms resulting in enhanced total antioxidant
capacity may play a role in modulating responses to O3.
Collectively these older and more recent studies provide insight into mechanisms
which may contribute to enhanced responses of asthmatic and atopic individuals
following O3 exposure. Greater airways inflammation and/or greater bronchial
reactivity have been demonstrated in asthmatics compared to non-asthmatics. This
pre-existing inflammation and altered baseline bronchial reactivity may contribute to
the enhanced bronchoconstriction seen in asthmatics exposed to O3. Furthermore,
O3-induced inflammation may contribute to O3-mediated AHR. An enhanced
neutrophilic response to O3 has been demonstrated in some asthmatics. A recent
study in humans provided evidence for differences in innate immune responses
related to TLR4 signaling between asthmatics and healthy subjects. Animal studies
have demonstrated a role for eosinophil-derived proteins in mediating the effects of
O3. Since airways eosinophilia occurs in both allergic humans and allergic animal
models, this pathway may underlie the exacerbation of allergic asthma by O3.
In addition, differences have been noted in epithelial cytokine expression in
bronchial biopsy samples of healthy and asthmatic subjects. A Th2 phenotype,
indicative of adaptive immune system activation and enhanced allergic responses,
was observed before O3 exposure and was increased by O3 exposure in asthmatics.
These findings support links between innate and adaptive immunity and sensitivity to
O3-mediated effects in asthmatics and allergic airways disease.
In addition to asthma and allergic diseases, obesity may alter responses to O3. While
O3 is a trigger for asthma, obesity is a known risk factor for asthma (Shore. 2007).
The relationship between obesity and asthma is not well understood but recent
investigations have focused on alterations in endocrine function of adipose tissue in
obesity. It is thought that the increases in serum levels of factors produced by
adipocytes (i.e., adipokines), such as cytokines, chemokines, soluble cytokine
receptors and energy regulating hormones, may contribute to the relationship
between obesity and asthma. Some of these same mechanisms may be relevant to
insulin resistant states such as metabolic syndrome.
In a re analysis of the data of Hazucha et al. (2003). increasing body mass index in
young women was associated with increased O3 responsiveness, as measured by
spirometry following a 1.5-hour exposure to 420 ppb O3 while exercising at a
moderate level (Bennett et al.. 2007). In several mouse models of obesity, airways
were found to be innately more hyperresponsive and responded more vigorously to
acute O3 exposure than lean controls (Shore. 2007). Pulmonary inflammatory and
injury in response to O3 were also enhanced (Shore. 2007). It was postulated that
oxidative stress resulting from obesity-related hyperglycemia could account for these
effects (Shore. 2007). However, responses to O3 in the different mouse models are
somewhat variable and depend on whether exposures are acute or subacute. For
example, diet-induced obesity augmented inflammation and injury, as measured by
BALF markers, and enhanced AHR in mice exposed acutely to O3 (2 ppm, 3 hours)
5-62
-------
(Johnston et al., 2008). In contrast, the inflammatory response following sub-acute
exposure to O3 was dampened by obesity in a different mouse model (0.3 ppm,
72 hours) (Shore et al., 2009). It is not known whether differences in responsiveness
to O3 are due to differences in lung development in genetically-modified animals
which result in smaller lungs and thus to differences in inhaled dose because of the
altered body mass to lung size ratio.
5.4.2.3 Nutritional Status
A further consideration is the compromised status of ELF antioxidants in nutritional
deficiencies. Thus, many investigations have focused on antioxidant deficiency and
supplementation as modulators of O3-mediated effects. One study in mice found that
vitamin A deficiency enhanced lung injury induced by exposure to 0.3 ppm O3 for
72 hours (Paquette et al., 1996). Ascorbate deficiency was shown to increase the
effects of acute (0.5-1 ppm for 4 hours), but not subacute (0.2-0.8 ppm for 7 days),
O3 exposure in guinea pigs (Kodavanti et al., 1995; Slade et al., 1989).
Supplementation with AH2 and a-TOH was protective in healthy adults who were on
an AH2-deficient diet and exposed to 400 ppb O3 for 2 hours while exercising at a
moderate level (Samet et al., 2001). In this study, the protective effect consisted of a
smaller reduction in FEVi following O3 exposure (Samet et al., 2001). However the
inflammatory response (influx of neutrophils and levels of IL-6) measured in BALF
1 hour after O3 exposure was not different between supplemented and non-
supplemented subjects (Samet et al., 2001). Other investigators found that AH2 and
a-TOH supplementation failed to ameliorate the pulmonary function decrements or
airways neutrophilia observed in humans exposed to 200 ppb O3 for 2 hours while
exercising at a moderate level (Mudwav et al.. 2006). It was suggested that
supplementation may be ineffective in the absence of antioxidant deficiency
(Mudwav et al.. 2006).
In asthmatic adults, these same dietary antioxidants reduced O3-induced bronchial
hyperresponsiveness (120 ppb, 45 min, light exercise) (Trenga et al., 2001).
Furthermore, supplementation with AH2 and a-tocopherol protected against
pulmonary function decrements and nasal inflammatory responses which were
associated with high levels of ambient O3 in asthmatic children living in
Mexico City, Mexico (Sienra-Monge et al., 2004; Romieu et al., 2002). Similarly,
supplementation with ascorbate, a-tocopherol and (3-carotene improved pulmonary
function in Mexico City street workers (Romieu et al., 1998b).
Protective effects of supplementation with a-tocopherol alone have not been
observed in humans experimentally exposed to O3 (Mudwav and Kelly. 2000).
Alpha-TOH supplementation also failed to protect against O3-induced effects in
animal models of allergic rhinosinusitis and lower airways allergic inflammation
(rats, 1 ppm O3 for 2 days) (Wagner et al.. 2007). However, protection in these same
animal models was reported using y-TOH supplementation (Wagner et al.. 2009;
Wagner et al.. 2007). Whether or not this effect was due to enhanced antioxidant
5-63
-------
status or to activated signaling pathways is not known. Other investigators found that
a-TOH deficiency led to an increase in liver lipid peroxidation following O3
exposure (rats, 0.3 ppm 3 hours/day for 7 months) (Sato et al.. 1980) and a drop in
liver a-TOH levels following O3 exposure (mice, 0.5 ppm, 6 hours/day for 3 days)
(Vasu et al., 2010). A recent study used a-TOH transfer protein null mice as a model
of a-TOH deficiency and demonstrated an altered adaptive response of the lung
genome to O3 exposure (Vasu et al.. 2010). Taken together, these studies provide
evidence that the tocopherol system modulates O3-induced responses.
5.4.2.4 Lifestage
Responses to O3 are modulated by factors associated with lifestage. On one end of
the lifestage spectrum is aging. The spirometric response to O3 appears to be lost in
humans as they age, as was demonstrated in two studies involving exposures of
human subjects exercising at levels ranging from light to heavy to 420-450 ppb O3
for 1.5-2 hours (Hazucha et al., 2003; Drechsler-Parks, 1995). In mice, physiological
responses to O3 (600 ppb, 2 hours) were diminished with age (Hamade et al., 2010).
Mechanisms accounting for this effect have not been well-studied but could include
altered number and sensitivity of receptors, altered signaling pathways involved in
neural reflexes or compromised status of ELF antioxidants.
On the other side of the lifestage spectrum is pre/postnatal development. Critical
windows of development during the pre/postnatal period are associated with an
enhanced sensitivity to environmental toxicants. Adverse birth outcomes and
developmental disorders may occur as a result (Section 7.4).
Adverse birth outcomes may result from stressors which impact transplacental
oxygen and nutrient transport by a variety of mechanisms including oxidative stress,
placental inflammation and placental vascular dysfunction (Kannan et al.. 2006).
These mechanisms may be linked since oxidative/nitrosative stress is reported to
cause vascular dysfunction in the placenta (Myatt et al.. 2000). As described earlier
in this chapter and in Section 7.4. systemic inflammation and oxidative/nitrosative
stress and modification of innate and adaptive immunity are key events underlying
the health effects of O3 and as such they may contribute to adverse birth outcomes.
An animal toxicology study showing that exposure to 2 ppm O3 led to anorexia
(Kavlock et al.. 1979) (see Section 7.4.2) in exposed rat dams provide an additional
mechanism by which O3 exposure could lead to diminished transplacental nutrient
transport. Disturbances of the pituitary-adrenocortico-placental system (Ritz et al..
2000) may also impact normal intrauterine growth and development. Further,
restricted fetal growth may result from pro-inflammatory cytokines which limit
trophoblast invasion during the early stages of pregnancy (Hansen et al.. 2008).
Direct effects on maternal health, such as risk of infection, and on fetal health, such
as DNA damage, have also been proposed as mechanisms underlying adverse birth
outcomes (Ritz et al.. 2000). In addition to restricted fetal growth, preterm birth may
contribute to adverse birth outcomes. Preterm birth may result from the development
5-64
-------
of premature contractions and/or premature rupture of membranes as well as from
disrupted implantation and placentation which results in suboptimal placental
function (Darrow et al., 2009; Ritz et al., 2000). Genetic mutations are thought to be
an important cause of placental abnormalities in the first trimester, while vascular
alterations may be the main cause of placental abnormalities in later trimesters
(Jalaludin et al.. 2007). Ozone-mediated systemic inflammation and oxidative
stress/nitrosative stress may possibly be related to these effects although there is no
firm evidence.
Enhanced sensitivity to environmental toxicants during critical windows of
development may also result in developmental disorders. For example, normal
migration and differentiation of neural crest cells are important for heart development
and are particularly sensitive to toxic insults (Ritz et al., 2002). Further, immune
dysregulation and related pathologies are known to be associated with pre/postnatal
environmental exposures (Dietert et al., 2010). Ozone exposure is associated with
developmental effects in several organ systems. These include the lung and immune
system (see below) and neurobehavioral changes which could reflect the effect of O3
on CNS plasticity or the hypothalamic-pituitary axis (Auten and Foster, 2011) (see
Section 7.4.9).
The majority of developmental effects due to O3 have been described for the
respiratory system (see Sections 7.2.3 and 7.4.8). Since its growth and development
take place during both the prenatal and early postnatal periods, both prenatal and
postnatal exposures to O3 have been studied. Maternal exposure to 0.4-1.2 ppm O3
during gestation resulted in developmental health effects in the RT of mice
(Sharkhuu et al.. 2011). Recent studies involving postnatal exposure to O3 have
focused on differences between developing and adult animals in antioxidant
defenses, respiratory physiology and sensitivity to cellular injury, and on
mechanisms, such as lung structural changes, antigen sensitization, interaction with
nitric oxide signaling, altered airway afferent innervation and loss of alveolar repair
capacity, by which early O3 exposure could lead to asthma pathogenesis or
exacerbations in later life (Auten and Foster, 2011).
An interesting set of studies conducted over the last 10 years in the infant rhesus
monkey has identified numerous O3-mediated perturbations in the developing lung
and immune system (Plopper et al.. 2007). These investigations were prompted by
the dramatic rise in the incidence of childhood asthma and focused on the possible
interaction of O3 and allergens in promoting remodeling of the epithelial -
mesenchymal trophic unit during postnatal development of the tracheobronchial
airway wall. In humans, airways growth during the 8-12 year period of postnatal
development is not well understood. Rhesus monkeys were used in these studies
because the branching pattern and distribution of airways in this model are more
similar to humans than those of rodents are to humans. In addition, a model of
allergic airways disease, which exhibits the main features of human asthma, had
already been established in the adult rhesus monkey. Studies in infant monkeys were
designed to determine whether repeated exposure to O3 altered postnatal growth and
development, and if so, whether such effects were reversible. In addition, exposure to
5-65
-------
O3 was evaluated for its potential to increase the development of allergic airways
disease. Exposures were to cyclic episodic O3 over 5 months, which involved 5
biweekly cycles of alternating filtered air and O3 - 9 consecutive days of filtered air
and 5 consecutive days of 0.5 ppm O3, 8 h/day - and to house dust mite allergen
(HDMA) for 2 hours per day for 3 days on the last 3 days of O3 exposure followed
by 11 days of filtered air.
Key findings were numerous. First, baseline airway resistance and AHR in the infant
monkeys were dramatically increased by combined exposure to both HDMA and O3
(Joad et al., 2006; Schelegle et al., 2003). Secondly, O3 exposure led to a large
increase in BAL eosinophils (Schelegle et al., 2003) while HDMA exposure led to a
large increase of eosinophils in airways tissue (Joad et al., 2006; Schelegle et al.,
2003). Thirdly, the growth pattern of distal airways was changed to a large extent by
exposure to O3 alone and in combination with HDMA. More specifically, longer and
narrower airways resulted and the number of conducting airway generations between
the trachea and the gas exchange area was decreased (Fanucchi et al., 2006). This
latter effect was not ameliorated by a recovery period of 6 months. Fourthly,
exposure to both HDMA and O3 altered the abundance and distribution of CD25+
lymphocytes in the airways (Miller et al., 2009). Lastly, several effects were seen at
the level of the epithelial mesenchymal trophic unit in response to O3. These include
altered organization of the airways epithelium (Schelegle et al., 2003), increased
abundance of mucous goblet cells (Schelegle et al., 2003), disruption of the basement
membrane zone (Evans et al., 2003), reduced innervation (Larson et al., 2004),
increased neuroendocrine-like cells (Joad et al., 2006), and altered orientation and
abundance of smooth muscle bundles (Plopper et al., 2007; Tran et al., 2004).
Six months of recovery in filtered air led to reversal of some but not all of these
effects OCaiekar et al.. 2007; Plopper et al.. 2007; Evans et al.. 2004). The authors
concluded that cyclic challenge of infant rhesus monkeys to allergen and O3 during
the postnatal period compromised airway growth and development and resulted in
changes which favor allergic airways responses and persistent effects on the immune
system (Plopper et al.. 2007). A more recent study in this same infant rhesus monkey
model reported that early life exposure to O3 resulted in decreased total peripheral
blood leukocyte numbers and increased blood eosinophils as well as persistent
effects on pulmonary and systemic innate immunity (Maniar-Hew et al.. 2011).
Furthermore, the effect of cyclic episodic O3 exposure on nasal airways was studied
in the infant rhesus monkey model. The three-dimensional detail of the nasal
passages was analyzed for developing predictive dosimetry models and exposure-
dose-response relationships (Carey et al., 2007). The authors reported that the
relative amounts of the five epithelial cell types in the nasal airways of monkeys
remained consistent between infancy and adulthood [comparing to (Gross et al.,
1987; Gross et al., 1982)1. Cyclic episodic O3 exposure (as described in the previous
paragraphs) resulted in 50-80% decreases in epithelial thickness and epithelial cell
volume of the ciliated respiratory and transitional epithelium, confirming that these
cell types in the nasal cavity were the most sensitive to O3 exposure. The character
and location of nasal lesions resulting from O3 exposure were similar in the infant
monkeys and adult monkeys similarly exposed. However, the nasal epithelium of
5-66
-------
infant monkeys did not undergo nasal airway epithelial remodeling or adaptation
which occurs in adult animals following O3-mediated injury and which may protect
against subsequent O3 challenge. The authors suggested that infant monkeys may be
prone to developing persistent necrotizing rhinitis following episodic longer-term
exposures.
5.4.2.5 Attenuation of Responses
Repeated daily exposure to O3 often results in a reduction in the degree of a
response, i.e., an attenuation of response. This phenomenon may reflect
compensatory mechanisms and is not necessarily beneficial. Furthermore, there is
variability among the different O3-related endpoints in terms of response attenuation,
as will be described below. As a result, attenuation of some responses occurs
concomitantly with the exacerbation of others.
In responsive individuals, a striking degree of attenuation of the FEVi response
occurred following repeated daily exposures to O3. Generally, the young O3
responder was no longer responsive on the fourth or fifth day of consecutive daily O3
exposure (200-500 ppb O3 for 2-4 hours with light to heavy levels of exercise) and
required days to weeks of nonexposure in order for the subject to regain O3
responsiveness (Christian et al.. 1998: Devlin et al.. 1997: Linn et al.. 1982b:
Horvath et al.. 1981: Hackney et al.. 1977b). This phenomena has been reported for
both lung function and symptoms such as upper airway irritation, nonproductive
cough, substernal discomfort and pain upon deep inspiration (Linn et al.. 1982b:
Horvath et al.. 1981: Hackney et al.. 1977b). Repeated daily exposures also led to an
attenuation of the sRaw response in moderately exercising human subj ects exposed
for 4 hours to 200 ppb O3 (Christian et al., 1998) and to a dampened AHR response
compared with a single day exposure in light exercising human subjects exposed for
2 hours to 400 ppb O3 (Dimeo et al., 1981). However, one group reported persistent
small airway dysfunction despite attenuation of the FEVi response on the third day
of consecutive O3 exposure (250 ppb, 2 hours, with moderate exercise) (Frank et al.,
2001).
Studies in rodents also indicated an attenuation of the physiologic response measured
by breathing patterns and tidal volume following five consecutive days of exposure
to 0.35-1 ppm O3 for 2.25 hours (Tepper et al.. 1989). Attenuation of O3-induced
brady cardie responses, which also result from activation of neural reflexes, has been
reported in rodents (0.5-0.6 ppm O3, 2-6 h/day, 3-5 days (Hamade and Tankersley.
2009: Watkinson et al.. 2001).
Multi-day exposure to O3 has been found to decrease some markers of inflammation
compared with a single day exposure (Christian et al., 1998: Devlin et al., 1997). For
example, in human subjects exposed for 4 hours to 200 ppb O3 during moderate
exercise, decreased numbers of BAL neutrophils and decreased levels of BALF
fibronectin and IL-6 were observed after 4 days of consecutive exposure compared
with responses after 1 day (Christian et al., 1998). Results indicated an attenuation of
5-67
-------
the inflammatory response in both proximal airways and distal lung. However
markers of injury, such as lactate dehydrogenase (LDH) and protein in the BALF,
were not attenuated in this study (Christian et al.. 1998). Other investigators found
that repeated O3 exposure (200 ppb O3 for 4 hours on 4 consecutive days with light
exercise) resulted in increased numbers of neutrophils in bronchial mucosal biopsies
despite decreased BAL neutrophilia (Torres et al.. 2000). Other markers of
inflammation, including BALF protein and IL-6 remained elevated following the
multi-day exposure (Jorres et al.. 2000).
In rats, the increases in BALF levels of proteins, fibronectin, IL-6 and inflammatory
cells observed after one day of exposure to 0.4 ppm O3 for 12 hours were no longer
observed after 5 consecutive days of exposure (Van Bree et al., 2002). A separate
study in rats exposed to 0.35-1 ppm O3 for 2.25 hours for 5 consecutive days
demonstrated a lack of attenuation of the increase in BALF protein, persistence of
macrophages in the centriacinar region and histological evidence of progressive
tissue injury (Tepper et al., 1989). Findings that injury, measured by BALF markers
or by histopathology, persist in the absence of BAL neutrophila or pulmonary
function decrements suggested that repeated exposure to O3 may have serious long-
term consequences such as airway remodeling. In particular, the small airways were
identified as a site where cumulative injury may occur (Frank et al., 2001).
Some studies examined the recovery of responses which were attenuated by repeated
O3 exposure. In a study of humans undergoing heavy exercise who were exposed for
2 hours to 400 ppb O3 for five consecutive days (Devlin et al.. 1997). recovery of the
inflammatory responses which were diminished by repeated exposure required
10-20 days following the exposure (Devlin et al.. 1997). In an animal study
conducted in parallel (Van Bree et al.. 2002). full susceptibility to O3 challenge
following exposure to O3 for five consecutive days required 15-20 days recovery.
Several mechanisms have been postulated to explain the attenuation of some
responses observed in human subjects and animal models following repeated
exposure to O3. First, the upregulation of antioxidant defenses (or conversely, a
decrease in critical O3-reactive substrates) may protect against O3-mediated effects.
Increases in antioxidant content of the BALF have been demonstrated in rats exposed
to 0.25 and 0.5 ppm O3 for several hours on consecutive days (Devlin et al.. 1997:
Wiester et al.. 1996b: Tepper et al.. 1989). Second, IL-6 was demonstrated to be an
important mediator of attenuation in rats exposed to 0.5 ppm for 4 hours on two
consecutive days (Mckinney et al.. 1998). Third, a protective role for increases in
mucus producing cells and mucus concentrations in the airways has been proposed
(Devlin et al.. 1997). Fourth, epithelial hyperplasia or metaplasia may decrease
further effects due to subsequent O3 challenge (Carey et al.. 2007: Harkemaet al..
1987a: Harkema et al.. 1987b). These morphologic changes have been observed in
nasal and lower airways in monkeys exposed chronically to 0.15-0.5 ppm O3 and
reflect a persistent change in epithelial architecture which may lead to other
long-term sequelae. Although there is some evidence to support these possibilities,
there is no consensus on mechanisms underlying response attenuation. Recent studies
demonstrating that O3 activates TRP receptors suggest that modulation of TRP
5-68
-------
receptor number or sensitivity by repeated O3 exposures may also contribute to the
attenuation of responses.
In summary, the attenuation of pulmonary function responses by repeated exposure
to O3 has been linked to exacerbation of O3-mediated injury. Enhanced exposure to
O3 due to a dampening of the O3-mediated truncation of inspiration may be one
factor which contributes to this relationship.
5.4.2.6 Co-exposures with Particulate Matter
Numerous studies have investigated the effects of co-exposure to O3 and PM because
of the prevalence of these pollutants in ambient air. Results are highly variable and
depend on whether exposures are simultaneous or sequential, the type of PM
employed and the endpoint examined. Additive and interactive effects have been
demonstrated. For example, simultaneous exposure to O3 (120 ppb for 2 hours at
rest) and concentrated ambient particles (CAPs) in human subjects resulted in a
diminished systemic IL-6 response compared with exposure to CAPs alone (Urch_et
al.. 2010). However, exposure to O3 alone did not alter blood IL-6 levels (Urch et al..
2010). The authors provided evidence that O3 mediated a switch to shallow breathing
which may have accounted for the observed antagonism (Urch et al.. 2010). Further,
simultaneous exposure to O3 (114 ppb for 2 hours at rest) and CAPs but not exposure
to either alone, resulted in increased diastolic blood pressure in human subjects
(Fakhri et al.. 2009). Mechanisms underlying this potentiation of response were not
explored. In some strains of mice, pre-exposure to O3 (0.5 ppm for 2 hours)
modulated the effects of carbon black PM on heart rate, HRV and breathing patterns
(Hamade and Tankersley. 2009). Another recent study in mice demonstrated that
treatment with carbon nanotubes followed 12 hours later by O3 exposure (0.5 ppm
for 3 hours) resulted in a dampening of some of the pulmonary effects of carbon
nanotubes measured as markers of inflammation and injury in the BALF (Han et al.,
2008). Further, Harkema and Wagner (2005) found that epithelial and inflammatory
responses in the airways of rats were enhanced by co-exposure to O3 (0.5 ppm for
3 days) and LPS (used as a model of biogenic PM) or to O3 (1 ppm for 2 days) and
OVA (used as a model of an aero allergen). Lastly, a recent study demonstrated that
maternal exposure to particulate matter (PM) resulted in augmented lung
inflammation, airway epithelial mucous metaplasia and AHR in young mice exposed
chronically and intermittently to 1 ppm O3 (Auten et al., 2009).
In summary, many of the demonstrated responses to co-exposure were more than
additive. These findings are hard to interpret but demonstrate the complexity of
responses following combined exposure to PM and O3.
5-69
-------
5.4.2.7 Summary
Collectively, these earlier and more recent studies provide some evidence for
mechanisms that may underlie the variability in responsiveness seen among
individuals (Figure 5-9). Certain functional genetic polymorphisms, pre-existing
conditions and diseases, nutritional status, lifestage and co-exposures contribute to
altered risk of O3-induced effects. Attenuation of responses may also be important,
but it is incompletely understood, both in terms of the pathways involved and the
resulting consequences.
Dosimetric factors
Nutritional status
Lifestage
Attenuation factors
Co-exposures
Ozone + Respiratory Tract
Gene-environment interactions
Pre-existing diseases/conditions
COPD/smoking status
Asthma/allergic airways disease
Obesity/metabolic syndrome
\
Formation of secondary oxidation products
Activation
of neural
reflexes
Initiation of
inflammation
\7
Sensitization
of bronchial
smooth muscle
Systemic inflammation and
oxidative/nitrosative stress
\
Respiratory System Effects
Extrapulmonary Effects
Obesity/
Metabolic Stress
Lifestage
Attenuation
factors
Figure 5-9 Some factors, illustrated in yellow, that likely contribute to the
interindividual variability in responses resulting from inhalation
ofO3.
5-70
-------
5.5 Species Homology and Interspecies Sensitivity
The previous O3 AQCDs discussed the homology of responses in animals and
humans exposed to O3 and the interspecies differences that may affect these
responses and concluded that the acute and chronic functional responses of
laboratory animals to O3 appear qualitatively homologous to human responses. Thus,
animal studies can provide important data in determining cause-effect relationships
between exposure and health outcome that would be impossible to collect in human
studies. Furthermore, animal studies add to a better understanding of the full range of
potential O3-mediated effects.
Still, care must be taken when comparing quantitative dose-response relationships in
animal models to humans due to obvious interspecies differences. This section will
qualitatively describe basic concepts in species homology concerning both dose and
response to O3 exposure. Overall, there have been few new publications examining
interspecies differences in dosimetry and response to O3 since the last AQCD. These
studies do not overtly change the conclusions discussed in the previous document.
5.5.1 Interspecies Dosimetry
As discussed in Section 5.2.1. O3 uptake depends on complex interactions between
RT morphology, breathing route, rate, and depth, physicochemical properties of the
gas, physical processes of gas transport, as well as the physical and chemical
properties of the ELF and tissue layers. Understanding differences in these variables
between humans and experimental animals is important to interpreting delivered
doses in animal and human toxicology studies.
Physiological and anatomical differences exist between experimental species.
The structure of the URT is vastly different between rodents and humans but scales
according to body mass. The difference in the cross-sectional shape and size of the
nasal passages affects bulk airflow patterns, particularly the shape of major airflow
streams. The nasal epithelium is lined by squamous, respiratory, transitional, or
olfactory cells, depending on location. The differences in airflow patterns in the URT
mean that not all nasal surfaces and cell types receive the same exposure to inhaled
O3 leading to differences in local absorption and potential for site-specific tissue
damage. The morphology of the LRT also varies within and among species. Rats and
mice do not possess respiratory bronchioles; however, these structures are present in
humans, dogs, ferrets, cats, and monkeys. Respiratory bronchioles are abbreviated in
hamsters, guinea pigs, sheep, and pigs. The branching structure of the ciliated
bronchi and bronchioles also differs between species from being a rather symmetric
and dichotomous branching network of airways in humans and primates to a more
monopodial branching network in other mammals. In addition, rodents have fewer
terminal bronchioles due to a smaller lung size compared to humans or canines
(McBride. 1992). The cellular composition in the pulmonary region is similar across
mammalian species; at least 95% of the alveolar epithelial tissue is composed of
5-71
-------
Type I cells. However, considerable differences exist between species in the number
and type of cells in the TB airways. Differences also exist in breathing route and rate.
Primates are oronasal breathers, while rodents are obligate nasal breathers. Past
studies of the effect of body size on resting oxygen consumption also suggest that
rodents inhale more volume of air per lung mass than primates. These distinctions as
well as differences in nasal structure between primates and rodents affect the amount
of O3 uptake.
As O3 absorption and reactivity relies on ELF antioxidant substances (see
Section 5.2.3), variability in antioxidant concentrations and metabolism between
species may affect dose and O3-induced health outcomes. The thickness of the ELF
in the TB airways varies among species. Mercer et al. (1992) found that the human
ELF thickness in bronchi and bronchioles was 6.9 and 1.8 |^m, respectively,
compared to 2.6 and 1.9 j^m for the same locations in the rat. Guinea pigs and mice
have a lower basal activity of GSH transferase and GSH peroxidase, and lower a-
TOH levels in the lung compared to rats (Ichinose et al., 1988; Sagai et al., 1987).
Nasal lavage fluid analysis shows that humans have a higher proportion of their nasal
antioxidants as UA and low levels of AH2 whereas mice, rats, or guinea pigs have
high levels of AH2 and undetectable levels of UA. GSH is not detected in the nasal
lavage fluid of most of these species, but is present in monkey nasal lavage fluid.
Guinea pigs and rats have a higher antioxidant to protein ratio in nasal lavage fluid
and BALF than humans (Hatch, 1992). The BALF profile differs from the nasal
lavage fluid. Humans have a higher proportion of GSH and less AH2 making up their
BALF content compared to the guinea pigs and rats (Slade et al., 1993; Hatch, 1992).
Similar to the nose, rats have the highest antioxidant to protein mass ratio found in
BALF (Slade et al., 1993). Antioxidant defenses also vary with age (Servais et al.,
2005) and exposure history (Duanetal.. 1996). Duan et al. (1996); Duan et al. (1993)
reported that differences in antioxidant levels between species and lung regions did
not appear to be the primary factor in O3 induced tissue injury. However, a close
correlation between site-specific O3 dose, the degree of epithelial injury, and reduced
glutathione depletion was observed in monkeys (Plopper et al.. 1998).
Even with these differences, humans and animals are similar in the pattern of
regional O3 dose distribution. As discussed for humans in Section 5.2.2, O3 flux to
the air-liquid interface of the ELF slowly decreases distally in the TB region and then
rapidly decreases distally in the alveolar region (Miller et al., 1985). Modeled tissue
dose in the human RT, representing O3 flux to the liquid-tissue interface, is very low
in the trachea, increases to a maximum in the CAR, and then rapidly decreases
distally in the alveolar region (Figure 5-1 Oa). Similar patterns of O3 tissue dose
profiles normalized to inhaled O3 concentration were predicted for rat, guinea pig,
and rabbit [(Miller et al., 1988) (Figure 5-10a)1. Overton et al. (1987) modeled rat
and guinea pig O3 dose distribution and found that after comparing two different
morphometrically based anatomical models for each species, considerable difference
in predicted percent RT and alveolar region uptakes were observed. This was due to
the variability between the two anatomical models in airway path distance to the first
alveolated duct. As a result, the overall dose profile was similar between species
however the O3 uptake efficiency varied due to RT size and path length
5-72
-------
(Section 5.2.2). A similar pattern of O3 dose distribution was measured in monkeys
exposed to 0.4 and 1.0 ppm 18O3 (Hopper et al.. 1998) (Figure 5-10b). Less 18O was
measured in the trachea, proximal bronchus, and distal bronchus than was observed
in the respiratory bronchioles. Again indicating the highest concentration of O3 tissue
dose is localized to the CAR, which are the respiratory bronchioles in nonhuman
primates. In addition, the lowest 18O detected in the RT was in the parenchyma
(i.e., alveolar region), reflecting the rapid decrease in tissue O3 dose predicted by
models for the alveolar regions of humans and other animals.
Humans and animal models are similar in the pattern of regional O3 dose, but
absolute values differ. Hatch et al. (1994) reported that exercising humans exposed to
oxygen-18 labeled O3 (400 ppb) accumulated 4-5 times higher concentrations of O3
reaction product in BAL cells, surfactant and protein fractions compared to resting
rats similarly exposed (400 ppb) (Figure 5-11). The use of 18O was specifically
employed in an attempt to accurately measure O3 dose to BALF fractions and lung
tissue and was normalized to the dried mass of lavaged constituents. It was necessary
to expose resting rats to 2 ppm O3 to achieve the same BALF accumulation of 18O
reaction product that was observed in humans exposed to 400 ppb with intermittent
heavy exercise (VE ~60 L/min). The concentration of 18O reaction product in BALF
paralleled the accumulation of BALF protein and cellular effects of the O3 exposure
observed such that these responses to 2.0 ppm O3 were similar to those of the
400 ppb O3 in exercising humans. This suggests that animal data obtained in resting
conditions would underestimate the reaction of O3 with cells in the RT and
presumably the resultant risk of effect for humans. However these results should be
interpreted with caution given an important limitation in the 18O labeling technique
when used for interspecies comparisons. The reaction between O3 and some
reactants such as ascorbate produce 18O-labeled products that are lost during sample
processing. When levels of ascorbate or other such reactants vary between species,
this lost portion of the total 18O-reaction products will also vary, leading to
uncertainty in interspecies comparisons.
5-73
-------
a.
E
o"
•t-i
c
0)
E
CQ
I
-6
10
10",
10-.J
0>
o
CO
o
Q
0)
3
tO
co
H ,{•
10
10
b.
.-•' >•"
*' ' Human •
Rat
Guinea Pig —•
Rabbit —-_..
VT (mL) f (bpm)
800 15.0
1.98 66.0
2.63 60.9
13.20 38.8
(No absorption in the URT)
TB
I— Zone 0
Order 0
Generation 0
Generation 0
4
5
6
10
10
5
7
12
12
14
14
6 7
91011
151617
161718
12
19
20
13
21
22
14
23
23
Rabbit
Guinea Pig
Rat
Human
TRACHEA
PROXIMAL
BRONCHUS
DISTAL
BRONCHUS
RESPIRATORY PARENCHYMA
BRONCHIOLE
Note: Panel (a) presents the predicted tissue dose of O3 (as ug of O3 per cm2 of segment surface area per min, standardized to a
tracheal O3 value of 1 ug/m3) for various regions of the rabbit, guinea pig, rat, and human RT. TB = tracheobronchial region,
A = alveolar region. Panel (b) presents a comparison of excess 18O in the five regions of the TB airways of rhesus monkeys
exposed to O3 for 2h. *p <0.05 comparing the same O3 concentration across regions. **p <0.05 comparing different O3
concentrations in the same region.
Source: Panel (a) Miller et al. (1988).Sprinaer-Verlag (b) Plopperetal. (1998)
Figure 5-10 Humans and animals are similar in the regional pattern of Oz
tissue dose distribution.
5-74
-------
BAL Cells
BAL HSP
BAL HSS
Lavaged Lung
Exercising Human
(0.4 ppm, 2 hours)
Resting F-344 Rat
(0.4 ppm, 2 hours)
Resting F-344 Rat
(2.0 ppm, 2 hours)
Note: The excess 18O in each fraction is expressed relative to the dry weight of that fraction. Fractions assayed include cells, high
speed pellet (HSP), high speed supernatant (HSS), and lavaged lung homogenates.
Source: Hatch etal. (1994)
Figure 5-11 Oxygen-18 incorporation into different fractions of BALF from
humans and rats exposed to 0.4 and 2.0 ppm 18Os.
Recently, a quantitative comparison of O3 transport in the airways of rats, dogs, and
humans was conducted using a three-compartment airways model, based on upper
and lower airway casts and mathematical calculation for alveolar parameters (Tsujino
et al., 2005). This one-dimensional gas transport model examined how interspecies
anatomical and physiological differences affect intra-airway O3 concentrations and
the amount of gas absorbed. The morphological model consisted of cylindrical tubes
with constant volume and no airway branching patterns. Peak, real-time, and mean
O3 concentrations were higher in the upper and lower airways of humans compared
to rats and dogs, but lowest in the alveoli of humans. The amount of O3 absorbed
was lowest in humans when normalized by body weight. The intra-airway
concentration decreased distally in all species. Sensitivity analysis demonstrated that
5-75
-------
VT, fe, and upper and lower airways surface area had a statistically significant
impact on model results. The model is limited in that it did not account for chemical
reactions in the ELF or consider gas diffusion as a driving force for O3 transport.
Also, the model was run at a respiratory rate of 16/min simulating a resting
individual, however exercise may cause a further deviation from animal models as
was seen in Hatch et al. (1994).
Overall, animal models exhibit qualitatively similar patterns of O3 net and tissue
dose distribution with the largest tissue dose delivered to the CAR. However, due to
anatomical and biochemical RT differences the absolute values of O3 dose delivered
differs. Past results suggest that animal data obtained in resting conditions would
underestimate the O3 reactions with cells in the BALF and presumably the resultant
risk of effect for humans, especially for humans during exercise.
5.5.2 Interspecies Homology of Response
Biological response to O3 exposure broadly shows commonalities in many species.
Among rodents, non-human primates, and humans, for example, ample data suggest
that O3 induces oxidative stress, cell injury, upregulation of cytokines/chemokines,
inflammation, alterations in lung function, and disruption of normal lung growth and
development (see Chapters 6 and 7).
The effects related to early life exposures can differ appreciably across species due to
the maturation stage of the lung and immune systems at birth. Evidence from non-
human primate studies shows that early life O3 exposure disrupts lung development
producing physiologic perturbations that are similar to those observed in children
exposed to urban air pollution (Fanucchi et al., 2006; Joad et al., 2006). Studies of O3
effects on lung surface chemistry also show some degree of homology. Lipid
oxidation products specific to O3 reactions with unsaturated fatty acids have been
reported, for example, in lavage fluids from both rodents and humans (Frampton et
al.. 1999: Pryor et al.. 1996). In humans, the extent to which systemic effects occur is
less well studied; plasma indices of lipid oxidation such as isoprostanes unfortunately
do not pinpoint the compartment(s) where oxidative stress has transpired. That
oxi dative stress occurs systemically in both rodents and non-human primates
(Chuang et al.. 2009). nevertheless, suggests that it likely also occurs in humans.
Despite the overall similarities in responses to O3 among species, studies have
reported variability in the responsiveness to O3 between and within species, as well
as between endpoints. Rodents appear to have a slightly higher tachypneic response
to O3 and are less sensitive to changes in pulmonary function responses than humans
(U.S. EPA, 1996a). However, rats experience attenuation of pulmonary function and
tachypneic ventilatory responses, similar to humans (Wiester et al., 1996b). Hatch et
al. (1986) reported that guinea pigs were the most responsive to O3-induced
inflammatory cell and protein influx. Rabbits were the least responsive and rats,
hamsters, and mice were intermediate responders. Further analysis of this study by
Miller et al. (1988) found that the protein levels in BALF from guinea pigs increased
5-76
-------
more rapidly with predicted pulmonary tissue dose than in rats and rabbits. Alveolar
macrophages isolated from guinea pigs and humans mounted similar qualitative and
quantitative cytokine responses to in vitro O3 (0.1-1.0 ppm for 60 minutes) exposure
(Arsalane et al., 1995).
Also, because of their higher body surface to volume ratio, rodents can rapidly lower
body temperature during exposure leading to lowered O3 dose and toxicity
(Watkinsonetal.. 2003: Iwasaki et al.. 1998: Sladeetal.. 1997). In addition to
lowering the O3 dose to the lungs, this hypothermic response may cause: (1) lower
metabolic rate, (2) altered enzyme kinetics, and (3) altered membrane function.
The thermoregulatory mechanisms also may disrupt heart rate that may lead to: (1)
decreased cardiac output, (2) lowered blood pressure, and (3) decreased tissue
perfusion (Watkinson et al., 2003). These responses have not been observed in
humans except at very high exposures, thus further complicating extrapolation of
effects from animals to humans.
The degree to which O3 induces injury and inflammation responses appears to be
variable between species. However, the majority of those studies did not normalize
the response to the dose received to account for species differences in O3 absorption.
For example, Dormans et al. (1999) found that rats, mice, and guinea pigs all
exhibited O3-induced (0.2 - 0.4 ppm for 3-56 days) inflammation; however, guinea
pigs were the most responsive with respect to alveolar macrophage elicitation and
pulmonary cell density in the centriacinar region. Mice were the most responsive in
terms of bronchiolar epithelial hypertrophy and biochemical changes (e.g., LDH,
glutathione reductase, glucose-6-phosphate dehydrogenase activity), and had the
slowest recovery from O3 exposure. All species displayed increased collagen in the
ductal septa and large lamellar bodies in Type II pneumocytes at the longest
exposure and highest concentration; whereas this response occurred in the rat and
guinea pig at lower O3 levels (0.2 ppm) as well. Overall, the authors rated mice as
most responsive, followed by guinea pigs, then rats (Dormans et al., 1999). Rats were
also less responsive in terms of epithelial necrosis and inflammatory responses as a
result of O3 exposure (1.0 ppm for 8 hours) compared with monkeys and ferrets,
which manifested a similar response (Sterner-Kock et al., 2000). Results of this study
should be interpreted with caution since no dose metric was used to normalize the
total inhaled dose or local organ dose between species.
To further understand the genetic basis for age-dependent differential response to O3,
adult (15 week old) and neonatal (15-16 day old) mice from 8 genetically diverse
strains were examined for O3-induced (0.8 ppm for 5 hours) pulmonary injury and
lung inflammation (Vancza et al.. 2009). Ozone exposure increased
polymorphonuclear leukocytes (PMN) influx in all strains of neonatal mice tested,
but significantly greater PMNs occurred in neonatal compared to adult mice for only
some sensitive strains, suggesting a genetic background effect. This strain difference
was not due to differences in delivered dose of O3 to the lung, evidenced by 18O lung
enrichment. The sensitivity of strains for O3-induced increases in BALF protein and
PMNs was different for different strains of mice suggesting that genetic factors
contributed to heightened responses. Interestingly, adult mice accumulated more than
5-77
-------
twice the levels of 18O reaction product of O3 than corresponding strain neonates.
Thus, it appeared that the infant mice showed a 2-fold- to 3-fold higher response than
the adults when expressed relative to the accumulated O3 reaction product in their
lungs. The apparent decrease in delivered O3 dose in neonates could be a result of a
more rapid loss of body temperature in infant rodents incident to maternal separation
and chamber air flow.
In animal studies, inhaled O3 concentration and exposure history rarely reflect actual
human environmental exposures. Generally, very high exposure concentrations are
used to induce murine AHR, which in some human subjects is observed at far more
relevant concentrations. This calls into question whether the differences in airway
reactivity are simply a function of differential nasopharyngeal scrubbing or whether
the complexities encompassing a variety of contributory biological pathways show
species divergence. Furthermore, in non-human primates exposed during early life,
eosinophil trafficking occurs, which has not been observed in rodents (unless
sensitized) (Maniar-Hew et al., 2011). This response has been shown to be persistent
when O3 challenges are administered after a recovery period of >9 months during
which no exposure transpired.
Quantitative extrapolation is challenging due to a number of uncertainties.
Unfortunately, many input parameters needed to conduct quantitative extrapolations
across species have not been obtained or currently remain undefined. It is not clear
whether characterization of the ELF provides the information needed to compute a
profile of reaction products or whether environmentally relevant exposure has altered
the physicochemical interactions that occur within the RT surface compartment
(e.g., O3 diffusion through regions where the ELF is thin). That systemic effects have
been documented in both rodents and non-human primates leads to the question of
whether reaction products, cytokines/chemokines, or both enter the nasopharyngeal
or bronchial circulation, both of which show species-dependent differences (Chuang
et al.. 2009: Cole and Freeman. 2009).
In addition, the response to O3 insult across species and more recent health effects
such as immune system development are uncertain. Non-human primate studies have
shown hypo-responsiveness to endotoxin challenge as a consequence of exposure;
whether this occurs in rodents and humans is largely unknown (Maniar-Hew et al..
2011). In addition, structural changes (e.g., airways remodeling, fibrogenesis) might
differ appreciably across species. Moreover, whether the upper airways differentially
contribute to either distal lung or systemic impacts has not been explored.
Some outcomes (e.g., inflammation) support the conclusion of homologous
responses across species. However, factors such as age, exposure history, diet,
endogenous substrate generation and homeostatic regulation, the cellular machinery
that regulates inflammatory cell trafficking, responses to other environmental
challenges, and the precise chemical species (whether ELF or cell membrane-
derived) that account for exposure-related initiation of pathophysiologic sequelae
might differ across species, but the extent of species-specific contributing factors
remains unknown. Consequently, some level of uncertainty cannot be dismissed.
Nonetheless, if experimental animals show pathophysiological consequences of
5-78
-------
exposure, the overall weight of the toxicological evidence supports the likelihood
that qualitatively similar effects occur in humans given appropriate exposure
scenarios.
5.5.3 Summary
In summary, biological response to O3 exposure broadly shows commonalities in
many species and thus supports the use of animal studies in determining mechanistic
and cause-effect relationships and as supporting evidence that similar effects could
occur in humans if O3 exposure is sufficient. However, there is uncertainty regarding
the similarity of response to O3 across species for some recently described endpoints.
Differences exist between species in a number of factors that influence O3 dosimetry
and responses, such as RT anatomy, breathing patterns, and ELF antioxidant
concentrations and chemical species. While humans and animals are similar in the
pattern of regional O3 dose distribution, these differences will likely result in
differences in the absolute values of O3 dose delivered throughout the RT. Thus,
these considerations can impact quantitative comparison between species.
5.6 Chapter Summary
Ozone is a highly reactive gas and a powerful oxidant with a short half-life. Both O3
uptake and responses are dependent upon the formation of secondary reaction
products in the ELF; however more complex interactions occur. Total RT uptake in
humans at rest is 80-95% efficient and it is influenced by a number of factors
including RT morphology, breathing route, frequency, and volume, physicochemical
properties of the gas, physical processes of gas transport, as well as the physical and
chemical properties of the ELF and tissue layers. In fact, even though the average RT
dose may be at a level where health effects would not be predicted, local regions of
the RT may receive considerably higher than average doses due to RT
inhomogeneity and differences in the pathlengths, and therefore be at greater risk of
effects. About half of the O3 that will be absorbed from the airstream is removed in
the URT, which provides a defense against O3 entering the lungs. However, the local
dose to the URT tissue is site-specific and dependent on the nasal anatomy, nasal
fluid composition, and ventilation and airflow patterns of the nasal passageways.
The primary uptake site of O3 delivery to the LRT epithelium is believed to be the
CAR, however changes in a number of factors (e.g., physical activity) can alter the
distribution of O3 uptake in the RT. Ozone uptake is chemical reaction-dependent
and the substances present in the ELF appear in most cases to limit interaction of O3
with underlying tissues and to prevent penetration of O3 distally into the RT. Still,
reactions of O3 with soluble ELF components or possibly plasma membranes result
in distinct products, some of which are highly reactive and can injure and/or transmit
signals to RT cells.
5-79
-------
Thus, in addition to contributing to the driving force for O3 uptake, formation of
secondary oxidation products initiates pathways that provide the mechanistic basis
for health effects that are described in detail in Chapters 6 and 7 and that involve the
RT as well as extrapulmonary systems. These pathways include activation of neural
reflexes, initiation of inflammation, alteration of epithelial barrier function,
sensitization of bronchial smooth muscle, modification of innate and adaptive
immunity, airways remodeling, and systemic inflammation and oxidative/nitrosative
stress. With the exception of airways remodeling, these pathways have been
demonstrated in both animals and human subjects in response to the inhalation of O3.
Both dosimetric and mechanistic factors contribute to the understanding of
interindividual variability in responses to O3. This variability is influenced by
differences in RT volume and surface area, certain genetic polymorphisms,
pre-existing conditions and disease, nutritional status, lifestages, attenuation, and
co-exposures. Some of these factors also underlie differences in species homology
and sensitivity. Qualitatively, animal models exhibit similar patterns of O3 net and
tissue dose distribution with the largest tissue dose of O3 delivered to the CAR.
However, due to anatomical and biochemical RT differences, the absolute value of
delivered O3 dose differs, with animal data obtained in resting conditions
underestimating the dose to the RT and presumably the resultant risk of effect for
humans, especially humans during exercise. Even though interspecies differences can
complicate quantitative comparison between species, many short-term responses of
laboratory animals to O3 appear qualitatively homologous to those of the human.
Furthermore, animal studies add to a better understanding of the full range of
potential O3-mediated effects. Given the commonalities in many responses across
species, animal studies that observe O3-induced effects may be used as supporting
evidence that similar effects could occur in humans or in determining mechanistic
and cause-effect relationships if O3 exposure is sufficient.
5-80
-------
References
Abraham. WM; Delehunt. JC: Yerger. L; Marchette. B; Oliver. W. Jr. (1984). Changes in airway permeability
and responsiveness after exposure to ozone. Environ Res 34: 110-119.
Agarwal. A; Saleh. RA; Bedaiwv. MA. (2003). Role of reactive oxygen species in the pathophysiology of
human reproduction [Review]. Fertil Steril 79: 829-843.
Ahmad. S: Ahmad. A; McConville. G: Schneider. BK; Allen. CB; Manzer. R; Mason. RJ; White. CW. (2005).
Lung epithelial cells release ATP during ozone exposure: Signaling for cell survival. Free Radic Biol Med
39: 213-226. http://dx.doi.Org/10.1016/i.freeradbiomed.2005.03.009
Aibo. PI; Birmingham. NP; Lewandowski. R; Maddox. JF; Roth. RA; Ganev. PE; Wagner. JG: Harkema. JR.
(2010). Acute exposure to ozone exacerbates acetaminophen-induced liver injury in mice. Toxicol Sci 115:
267-285. http://dx.doi.org/10.1093/toxsci/kfq034
Al-Hegelan. M; Tighe. RM; Castillo. C: Hollingsworth. JW. (2011). Ambient ozone and pulmonary innate
immunity [Review]. Immunol Res 49: 173-191. http://dx.doi.org/10.1007/sl2026-010-8180-z
Alexis. N; Soukup. J; Nierkens. S: Becker. S. (200Ib). Association between airway hyperreactivity and
bronchial macrophage dysfunction in individuals with mild asthma. Am J Physiol Lung Cell Mol Physiol
280: L369-L375.
Alexis. N: Urch. B; Tarlo. S: Corey. P; Pengelly. D; O'Byrne. P; Silverman. F. (2000). Cyclooxygenase
metabolites play a different role in ozone-induced pulmonary function decline in asthmatics compared to
normals. Inhal Toxicol 12: 1205-1224.
Alexis. NE; Lav. JC: Hazucha. M; Harris. B; Hernandez. ML; Bromberg. PA; Kehrl. H; Diaz-Sanchez. D;
Kim. C: Devlin. RB; Peden. DB. (2010). Low-level ozone exposure induces airways inflammation and
modifies cell surface phenotypes in healthy humans. Inhal Toxicol 22: 593-600.
http://dx.doi.org/10.3109/08958371003596587
Alexis. NE; Zhou. H; Lav. JC: Harris. B; Hernandez. ML; Lu. TS: Bromberg. PA; Diaz-Sanchez. D; Devlin.
RB; Kleeberger. SR; Peden. DB. (2009). The glutathione-S-transferase Mu 1 null genotype modulates
ozone-induced airway inflammation in human subjects. J Allergy Clin Immunol 124: 1222-1228.
http://dx.doi.0rg/10.1016/i.iaci.2009.07.036
Alfaro. MF; Putney. L; Tarkington. BK; Hatch. GE; Hyde. DM; Schelegle. ES. (2004). Effect of rapid shallow
breathing on the distribution of ISO-labeled ozone reaction product in the respiratory tract of the rat. Inhal
Toxicol 16: 77-85.
Alfaro. MF; Walbv. WF; Adams. WC: Schelegle. ES. (2007). Breath condensate levels of 8-isoprostane and
leukotriene B4 after ozone inhalation are greater in sensitive versus nonsensitive subjects. Exp Lung Res
33: 115-133.
Araneda. S: Commin. L; Atlagich. M; Kitahama. K; Parraguez. VH; Pequignot. JM; Dalmaz. Y. (2008). VEGF
overexpression in the astroglial cells of rat brainstem following ozone exposure. Neurotoxicology 29: 920-
927. http://dx.doi.0rg/10.1016/i.neuro.2008.09.006
Aris. RM; Christian. D; Hearne. PQ: Kerr. K; Finkbeiner. WE; Balmes. JR. (1993). Ozone-induced airway
inflammation in human subjects as determined by airway lavage and biopsy. Am J Respir Crit Care Med
148: 1363-1372.
Arito. H; Uchiyama. I; Arakawa. H; Yokoyama. E. (1990). Ozone-induced bradycardia and arrhythmia and
their relation to sleep-wakefulness in rats. Toxicol Lett 52: 169-178. http://dx.doi.org/10.1016/0378-
4274(90)90151-6
5-81
-------
Arsalane. K; Gosset. P; Vanhee. D; Voisin. C: Hamid. Q: Tonnel. AB; Wallaert. B. (1995). Ozone stimulates
synthesis of inflammatory cytokines by alveolar macrophages in vitro. Am J Respir Cell Mol Biol 13: 60-
68.
Asplund. PT; Ben-Jebria. A; Rigas. ML; Ultman. JS. (1996). Longitudinal distribution of ozone absorption in
the lung: Effect of continuous inhalation exposure. Arch Environ Occup Health 51: 431-438.
Auten. RL; Foster. WM. (2011). Biochemical effects of ozone on asthma during postnatal development
[Review]. Biochim Biophys Acta 1810: 1114-1119. http://dx.doi.Org/10.1016/i.bbagen.2011.01.008
Auten. RL; Potts. EN; Mason. SN; Fischer. B; Huang. Y; Foster. WM. (2009). Maternal exposure to particulate
matter increases postnatal ozone-induced airway hyperreactivity in juvenile mice. Am J Respir Crit Care
Med 180: 1218-1226. http://dx.doi.org/10.1164/rccm.200901-0116OC
Awasthi. YC: Yang. Y; Tiwari. NK; Patrick. B; Sharma. A; Li. J: Awasthi. S. (2004). Regulation of 4-
hydroxynonenal-mediated signaling by gluathione S-transferases. Free Radic Biol Med 37: 607-619.
http://dx.doi.0rg/10.1016/i.freeradbiomed.2004.05.033
Backus. GS: Howden. R; Fostel. J: Bauer. AK; Cho. HY; Marzec. J: Peden. DB; Kleeberger. SR. (2010).
Protective role of interleukin-10 in ozone-induced pulmonary inflammation. Environ Health Perspect 118:
1721-1727. http://dx.doi.org/10.1289/ehp.1002182
Ballinger. CA; Cueto. R; Squadrito. G: Coffin. JF; Velsor. LW: Pryor. WA; Postlethwait. EM. (2005).
Antioxidant-mediated augmentation of ozone-induced membrane oxidation. Free Radic Biol Med 38: 515-
526. http://dx.doi.0rg/10.1016/j.freeradbiomed.2004.ll.009
Balmes. JR; Arts. RM; Chen. LL; Scannell. C: Tager. IB; Finkbeiner. W: Christian. D; Kelly. T; Hearne. PQ:
Ferrando. R: Welch. B. (1997). Effects of ozone on normal and potentially sensitive human subjects. Part I:
Airway inflammation and responsiveness to ozone in normal and asthmatic subjects (pp. 1-37; discussion
81-99). (ISSN 1041-5505). Boston, MA: Health Effects Institute.
Balmes. JR: Chen. LL; Scannell. C: Tager. I; Christian. D; Hearne. PQ: Kelly. T; Aris. RM. (1996). Ozone-
induced decrements in FEV1 and FVC do not correlate with measures of inflammation. Am J Respir Crit
Care Med 153: 904-909.
Bang. S: Kim. KY; Yoo. S: Kim. YG: Hwang. SW. (2007). Transient receptor potential Al mediates
acetaldehyde-evoked pain sensation. Eur J Neurosci 26: 2516-2523. http://dx.doi.org/10.1111/i. 1460-
9568.2007.05882.x
Basha. MA; Gross. KB; Gwizdala. CJ: Haidar. AH; Popovich. J. Jr. (1994). Bronchoalveolar lavage
neutrophilia in asthmatic and healthy volunteers after controlled exposure to ozone and filtered purified air.
Chest 106: 1757-1765.
Bastacky. J: Lee. CY; Goerke. J: Koushafar. H; Yager. D; Kenaga. L; Speed. TP; Chen. Y; Clements. JA.
(1995). Alveolar lining layer is thin and continuous: Low-temperature scanning electron microscopy of rat
lung. JAppl Physiol 79: 1615-1628.
Bates. ML; Brenza. TM; Ben-Jebria. A; Bascom. R; Ultman. JS. (2009). Longitudinal distribution of ozone
absorption in the lung: Comparison of cigarette smokers and nonsmokers. Toxicol Appl Pharmacol 236:
270-275.
Bauer. AK; Rondini. EA; Hummel. KA; Degraff. LM; Walker. C: Jedlicka. AE; Kleeberger. SR. (2011).
Identification of candidate genes downstream of TLR4 signaling after ozone exposure in mice: A role for
heat shock protein 70. Environ Health Perspect 119: 1091-1097. http://dx.doi.org/10.1289/ehp.1003326
Beckett. WS: McDonnell. WF; Horstman. DH; House. DE. (1985). Role of the parasympathetic nervous
system in acute lung response to ozone. JAppl Physiol 59: 1879-1885.
Becquemin. MM; Bertholon. JF; Bouchikhi. A; Malarbet. JL; Roy. M. (1999). Oronasal ventilation partitioning
in adults and children: Effect on aerosol deposition in airways. Radiat Prot Dosimetry 81: 221-228.
Belvisi. MG: Stretton. CD; Verleden. GM; Ledingham. SJ: Yacoub. MH; Barnes. PJ. (1992). Inhibition of
cholinergic neurotransmission in human airways by opioids. JAppl Physiol 72: 1096-1100.
5-82
-------
Bennett. W: Zeman. K; Jarabek. A. (2003). Nasal contribution to breathing with exercise: effect of race and
gender. J Appl Physiol 95: 497-503. http://dx.doi.org/10.1152/iapplphysiol.00718.2002
Bennett. WD; Hazucha. MJ; Folinsbee. LJ; Bromberg. PA; Kissling. GE; London. SJ. (2007). Acute
pulmonary function response to ozone in young adults as a function of body mass index. Inhal Toxicol 19:
1147-1154. http://dx.doi.org/10.1080/08958370701665475
Bennett. WD: Zeman. KL; Jarabek. AM. (2008). Nasal contribution to breathing and fine particle deposition in
children versus adults. J Toxicol Environ Health A 71: 227-237.
http://dx.doi.org/10.1080/15287390701598200
Bergamaschi. E; De Palma. G: Mozzoni. P; Vanni. S: Vettori. MV; Broeckaert. F; Bernard. A; Mutti. A.
(2001). Polymorphism of quinone-metabolizing enzymes and susceptibility to ozone-induced acute effects.
Am J Respir Crit Care Med 163: 1426-1431.
Bhalla. DK; Gupta. SK. (2000). Lung injury, inflammation, and inflammatory stimuli in rats exposed to ozone.
J Toxicol Environ Health 59: 211-228. http://dx.doi.org/10.1080/009841000156899
Block. ML; Calderon-Garciduenas. L. (2009). Air pollution: Mechanisms of neuroinflammation and CNS
disease [Review]. Trends Neurosci 32: 506-516. http://dx.doi.Org/10.1016/i.tins.2009.05.009
Blomberg. A; Mudwav. IS: Nordenhall. C: Hedenstrom. H; Kelly. FJ; Frew. AJ; Holgate. ST; Sandstrom. T.
(1999). Ozone-induced lung function decrements do not correlate with early airway inflammatory or
antioxidant responses. Eur Respir J 13: 1418-1428.
Bosson. J; Blomberg. A; Pourazar. J: Mudwav. IS: Frew. AJ: Kelly. FJ: Sandstrom. T. (2009). Early
suppression of NFkappaB and IL-8 in bronchial epithelium after ozone exposure in healthy human subjects.
Inhal Toxicol 21: 913-919. http://dx.doi.org/10.1080/08958370802657389
Bosson. J; Stenfors. N; Bucht. A; Helledav. R; Pourazar. J: Holgate. ST; Kelly. FJ: Sandstrom. T; Wilson. S:
Frew. AJ: Blomberg. A. (2003). Ozone-induced bronchial epithelial cytokine expression differs between
healthy and asthmatic subjects. Clin Exp Allergy 33: 777-782.
Broeckaert. F; Clippe. A; Wattiez. R; Falmagne. P; Bernard. A. (2003). Lung hyperpermeability, Clara-cell
secretory potein (CC16), and susceptibility to ozone of five inbred strains of mice. Inhal Toxicol 15: 1209-
1230.
Bush. ML; Asplund. PT; Miles. KA; Ben-Jebria. A; Ultman. JS. (1996). Longitudinal distribution of O3
absorption in the lung: gender differences and intersubject variability. J Appl Physiol 81: 1651-1657.
Bush. ML; Zhang. W: Ben-Jebria. A; Ultman. JS. (2001). Longitudinal distribution of ozone and chlorine in
the human respiratory tract: Simulation of nasal and oral breathing with the single-path diffusion model.
Toxicol Appl Pharmacol 173: 137-145. http://dx.doi.org/10.1006/taap.2001.9182
Caceres. AI; Brackmann. M; Elia. MD; Bessac. BF; del Camino. D; D'Amours. M; Witek. JS: Fanger. CM;
Chong. JA; Havward. NJ; Homer. RJ; Cohn. L; Huang. X; Moran. MM; Jordt. SE. (2009). A sensory
neuronal ion channel essential for airway inflammation and hyperreactivity in asthma. PNAS 106: 9099-
9104. http://dx.doi.org/10.1073/pnas.0900591106
Carey. SA; Minard. KR; Trease. LL; Wagner. JG: Garcia. GJ: Ballinger. CA; Kimbell. JS: Plopper. CG: Corlev.
RA; Postlethwait. EM; Harkema. JR. (2007). Three-dimensional mapping of ozone-induced injury in the
nasal airways of monkeys using magnetic resonance imaging and morphometric techniques. Toxicol Pathol
35: 27-40. http://dx.doi.org/10.1080/01926230601072343
Chang. LY; Huang. Y; Stockstill. BL; Graham. JA; Grose. EC: Menache. MG: Miller. FJ: Costa. PL; Crapo.
JD. (1992). Epithelial injury and interstitial fibrosis in the proximal alveolar regions of rats chronically
exposed to a simulated pattern of urban ambient ozone. Toxicol Appl Pharmacol 115: 241-252.
http://dx.doi.org/10.1016/0041-008X(92)90329-Q
Chen. C: Arjomandi. M; Balmes. J: Tager. I; N. H. (2007a). Effects of chronic and acute ozone exposure on
lipid peroxidation and antioxidant capacity in healthy young adults. Environ Health Perspect 115: 1732-
1737. http://dx.doi.org/10.1289/ehp.10294
5-83
-------
Chen. C: Arjomandi. M; Qin. H; Balmes. J: Tager. I; Holland. N. (2006a). Cytogenetic damage in buccal
epithelia and peripheral lymphocytes of young healthy individuals exposed to ozone. Mutagenesis 21: 131-
137. http://dx.doi.org/10.1093/mutage/gel007
Cho. HY; Hotchkiss. JA; Harkema. JR. (1999). Inflammatory and epithelial responses during the development
of ozone-induced mucous cell metaplasia in the nasal epithelium of rats. Toxicol Sci 51: 135-145.
Cho. HY; Kleeberger. SR. (2007). Genetic mechanisms of susceptibility to oxidative lung injury in mice. Free
Radic Biol Med 42: 433-445. http://dx.doi.Org/10.1016/i.freeradbiomed.2006.ll.021
Cho. HY; Morgan. PL; Bauer. AK; Kleeberger. SR. (2007). Signal transduction pathways of tumor necrosis
factor—mediated lung injury induced by ozone in mice. Am J Respir Crit Care Med 175: 829-839.
http://dx.doi.org/10.1164/rccm.200509-1527OC
Cho. HY; Zhang. LY; Kleeberger. SR. (2001). Ozone-induced lung inflammation and hyperreactivity are
mediated via tumor necrosis factor-alpha receptors. Am J Physiol 280: L537-L546.
Christian. PL; Chen. LL; Scannell. CH; Ferrando. RE; Welch. BS: Balmes. JR. (1998). Ozone-induced
inflammation is attenuated with multiday exposure. Am J Respir Crit Care Med 158: 532-537.
Chuang. GC: Yang. Z; Westbrook. DG: Pompilius. M; Ballinger. CA; White. RC: Krzywanski. DM;
Postlethwait. EM; Ballinger. SW. (2009). Pulmonary ozone exposure induces vascular dysfunction,
mitochondria! damage, and atherogenesis. Am J Physiol Lung Cell Mol Physiol 297: L209-L216.
http://dx.doi.org/10.1152/aiplung.00102.2009
Cohen-Hubal. EA; Kimbell. JS: Fedkiw. PS. (1996). Incorporation of nasal-lining mass-transfer resistance into
a CFD model for prediction of ozone dosimetry in the upper respiratory tract. Inhal Toxicol 8: 831-857.
Cole. MP; Freeman. BA. (2009). Promotion of cardiovascular disease by exposure to the air pollutant ozone
[Review]. Am J Physiol Lung Cell Mol Physiol 297: L209-L216.
http://dx.doi.org/10.1152/ajplung.00187.2009
Coleridge. HM; Coleridge. JCG: Ginzel. KH; Baker. DG: Banzett. RB; Morrison. MA. (1976). Stimulation of
'irritant' receptors and afferent C-fibers in the lungs by prostaglandins. Nature 264: 451-453.
Coleridge. JCG: Coleridge. HM; Schelegle. ES: Green. JF. (1993). Acute inhalation of ozone stimulates
bronchial C-fibers and rapidly adapting receptors in dogs. J Appl Physiol 74: 2345-2352.
Connor. LM; Ballinger. CA; Albrecht. TB; Postlethwait. EM. (2004). Interfacial phospholipids inhibit ozone
reactive absorption-mediated cytotoxicity in vitro. Am J Physiol 286: L1169-L1178.
http://dx.doi.org/10.1152/ajplung.00397.2003
Corradi. M; Alinovi. R; Goldoni. M; Vettori. M; Folesani. G: Mozzoni. P; Cavazzini. S: Bergamaschi. E;
Rossi. L; Mutti. A. (2002). Biomarkers of oxidative stress after controlled human exposure to ozone.
Toxicol Lett 134: 219-225.
Costa. PL; Schafrank. SN; Wehner. RW: Jellett E. (1985). Alveolar permeability to protein in rats
differentially susceptible to ozone. J Appl Toxicol 5: 182-186. http://dx.doi.org/10.1002/iat.2550050309
Cross. CE; Motchnik. PA; Bruener. BA; Jones. DA; Kaur. H; Ames. BN; Halliwell. B. (1992). Oxidative
damage to plasma constituents by ozone. FEES Lett 298: 269-272. http://dx.doi.org/10.1016/0014-
5793(92)80074-0
Dahl. M; Bauer. AK; Arredouani. M; Soininen. R; Trvggvason. K; Kleeberger. SR: Kobzik. L. (2007).
Protection against inhaled oxidants through scavenging of oxidized lipids by macrophage receptors
MARCO and SR-AI/II. J Clin Invest 117: 757-764. http://dx.doi.org/10.1172/JCI29968
Darrow. LA; Klein. M; Flanders. WD; Waller. LA; Correa. A; Marcus. M; Mulholland. JA; Russell. AG:
Tolbert. PE. (2009). Ambient air pollution and preterm birth: A time-series analysis. Epidemiology 20: 689-
698. http://dx.doi.org/10.1097/EDE.Ob013e3181a7128f
Delaunois. A; Segura. P; Montano. LM; Vargas. MH; Ansav. M; Gustin. P. (1998). Comparison of ozone-
induced effects on lung mechanics and hemodynamics in the rabbit. Toxicol Appl Pharmacol 150: 58-67.
5-84
-------
Dempsev. JA; Johnson. BD; Saupe. KW. (1990). Adaptations and limitations in the pulmonary system during
exercise [Review]. Chest 97: 81S-87S.
Dempsev. JA; McKenzie. DC: Haverkamp. HC: Eldridge. MW. (2008). Update in the understanding of
respiratory limitations to exercise performance in fit, active adults [Review]. Chest 134: 613-622.
http://dx.doi.org/10.1378/chest.07-2730
Devlin. RB; Duncan. KE; Jardim. M; Schmitt. MT; Rappold. AG: Diaz-Sanchez. D. (2012). Controlled
exposure of healthy young volunteers to ozone causes cardiovascular effects. Circulation 126: 104-111.
http://dx.doi.org/10.1161/CIRCULATIONAHA.112.094359
Devlin. RB; Folinsbee. LJ; Biscardi. F; Hatch. G: Becker. S: Madden. MC: Robbins. M; Koren. HS. (1997).
Inflammation and cell damage induced by repeated exposure of humans to ozone. Inhal Toxicol 9: 211-235.
Devlin. RB: Mcdonnell. WF; Mann. R; Becker. S: House. DE; Schreinemachers. D; Koren. HS. (1991).
Exposure of humans to ambient levels of ozone for 6.6 hours causes cellular and biochemical changes in
the lung. Am J Respir Cell Mol Biol 4: 72-81.
Diemer. T; Allen. JA: Hales. KH; Hales. DB. (2003). Reactive oxygen disrupts mitochondria in MA-10 tumor
Leydig cells and inhibits steroidogenic acute regulatory (StAR) protein and steroidogenesis. Endocrinology
144: 2882-2891. http://dx.doi.org/10.1210/en.2002-0090
Dietert. RR; DeWitt. JC: Germolec. PR; Zelikoff. JT. (2010). Breaking patterns of environmentally influenced
disease for health risk reduction: Immune perspectives [Review]. Environ Health Perspect 118: 1091-1099.
http://dx.doi.org/10.1289/ehp.1001971
Dimeo. MJ; Glenn. MG: Holtzman. MJ; Sheller. JR; Nadel. JA: Boushev. HA. (1981). Threshold concentration
of ozone causing an increase in bronchial reactivity in humans and adaptation with repeated exposures. Am
Rev Respir Dis 124: 245-248.
Dormans. JAM. A; Van Bree. L; Boere. AJF; Marra. M; Rombout. PJA. (1999). Interspecies differences in
time course of pulmonary toxicity following repeated exposure to ozone. Inhal Toxicol 11: 309-329.
http://dx.doi.org/10.1080/089583799197113
Dostert. C: Petrilli. V; Van Bruggen. R; Steele. C: Mossman. BT; Tschopp. J. (2008). Innate immune activation
through Nalp3 inflammasome sensing of asbestos and silica. Science 320: 674-677.
http://dx.doi.org/10.1126/science.1156995
Drechsler-Parks. DM. (1995). The dose-response relationship in older men exposed to ozone. Exp Gerontol 30:
65-75.
Duan. X; Buckpitt. AR; Pinkerton. KE; Ji. C: Plopper. CG. (1996). Ozone-induced alterations in glutathione in
lung subcompartments of rats and monkeys. Am J Respir Cell Mol Biol 14: 70-75.
Duan. X; Buckpitt. AR; Plopper. CG. (1993). Variation in antioxidant enzyme activities in anatomic
subcompartments within rat and rhesus monkey lung. Toxicol Appl Pharmacol 123: 73-82.
Emmons. K; Foster. WM. (1991). Smoking cessation and acute airway response to ozone. Arch Environ Occup
Health 46: 288-295. http://dx.doi.org/10.1080/00039896.1991.9934389
Enami. S: Hoffmann. MR; Colussi. AJ. (2008a). Acidity enhances the formation of a persistent ozonide at
aqueous ascorbate/ozone gas interfaces. PNAS 105: 7365-7369. http://dx.doi.org/10.1073/pnas.0710791105
Enami. S: Hoffmann. MR; Colussi. AJ. (2008b). Ozonolysis of uric acid at the air/water interface. J Phys
Chem B 112: 41534156. http://dx.doi.org/10.1021/jp712010k
Enami. S: Hoffmann. MR; Colussi. AJ. (2009a). How phenol and alpha-tocopherol react with ambient ozone at
gas/liquid interfaces. J Phys Chem A 113: 7002-7010. http://dx.doi.org/10.1021/ip901712k
Enami. S: Hoffmann. MR; Colussi. AJ. (2009b). Ozone oxidizes glutathione to a sulfonic acid. Chem Res
Toxicol 22: 35-40. http://dx.doi.org/10.1021/tx800298ilO.1021/tx800298i
5-85
-------
Enami. S: Hoffmann. MR; Colussi. AJ. (2009c). Simultaneous detection of cysteine sulfenate, sulfinate, and
sulfonate during cysteine interfacial ozonolysis [Letter]. J Phys Chem B 113: 9356-9358.
http://dx.doi.org/10.1021/ip904316n
Engel. LA. (1985). Intraregional gas mixing and distribution. In Gas mixing and distribution in the lung. New
York: Marcel Dekker.
Esther. CR; Peden. DB; Alexis. NE; Hernandez. ML. (2011). Airway purinergic responses in healthy, atopic
nonasthmatic, and atopic asthmatic subjects exposed to ozone. Inhal Toxicol 23: 324-330.
http://dx.doi.org/10.3109/08958378.2011.572096
Evans. MJ; Fanucchi. MV; Baker. GL; Van Winkle. LS: Pantle. LM; Nishio. SJ: Schelegle. ES: Gershwhin. LJ;
Miller. LA; Hyde. DM; Sannes. PL; Plopper. CG. (2003). Atypical development of the tracheal basement
membrane zone of infant rhesus monkeys exposed to ozone and allergen. Am J Physiol 285: L931-L939.
http://dx.doi.org/10.1152/ajplung.00175.2003
Evans. MJ: Fanucchi. MV: Baker. GL; Van Winkle. LS: Pantle. LM; Nishio. SJ: Schelegle. ES: Gershwin. LJ:
Miller. LA; Hyde. DM; Plopper. CG. (2004). The remodelled tracheal basement membrane zone of infant
rhesus monkeys after 6 months of recovery. Clin Exp Allergy 34: 1131-1136.
http://dx.doi.0rg/10.llll/i.1365-2222.2004.02004.x CEA2004
Fabbri. LM; Aizawa. H; O'Byrne. PM; Bethel. RA: Walters. EH; Holtzman. MJ: Nadel. JA. (1985). An anti-
inflammatory drug (BW755C) inhibits airway hyperresponsiveness induced by ozone in dogs. J Allergy
Clin Immunol 76: 162-166. http://dx.doi.org/10.1016/0091-6749(85)90695-5
Fakhri. AA; Ilic. LM; Wellenius. GA; Urch. B; Silverman. F; Gold. PR; Mittleman. MA. (2009). Autonomic
effects of controlled fine paniculate exposure in young healthy adults: Effect modification by ozone.
Environ Health Perspect 117: 1287-1292. http://dx.doi.org/10.1289/ehp.0900541
Fanucchi. MV: Plopper. CG: Evans. MJ: Hyde. DM; Van Winkle. LS: Gershwin. LJ: Schelegle. ES. (2006).
Cyclic exposure to ozone alters distal airway development in infant rhesus monkeys. Am J Physiol Lung
Cell Mol Physiol 291: L644-L650. http://dx.doi.org/10.1152/aiplung.00027.2006
Feng. R; He. W: Ochi. H; Castranova. V. (2006). Ozone exposure impairs antigen-specific immunity but
activates IL-7-induced proliferation of CD4-CD8- thymocytes in BALB/c mice. J Toxicol Environ Health
A69: 1511-1526. http://dx.doi.org/10.1080/15287390500468696
Foster. WM; Brown. RH; Macri. K; Mitchell. CS. (2000). Bronchial reactivity of healthy subjects: 18-20 h
postexposure to ozone. J Appl Physiol 89: 1804-1810.
Foster. WM; Freed. AN. (1999). Regional clearance of solute from peripheral airway epithelia: Recovery after
sublobar exposure to ozone. J Appl Physiol 86: 641-646.
Foster. WM: Stetkiewicz. PT. (1996). Regional clearance of solute from the respiratory epithelia: 18-20 h
postexposure to ozone. J Appl Physiol 81: 1143-1149.
Foster. WM; Wills-Karp. M; Tankerslev. CG: Chen. X; Paquette. NC. (1996). Bloodborne markers in humans
during multiday exposure to ozone. J Appl Physiol 81: 794-800.
Frampton. MW: Morrow. PE; Torres. A; Cox. C: Voter. KZ: Utell. MJ: Gibb. FR; Speers. DM. (1997a). Ozone
responsiveness in smokers and nonsmokers. Am J Respir Crit Care Med 155: 116-121.
Frampton. MW: Morrow. PE; Torres. A; Voter. KZ: Whitin. JC: Cox. C: Speers. DM; Tsai. Y; Utell. MJ.
(1997b). Effects of ozone on normal and potentially sensitive human subjects: Part II. Airway inflammation
and responsiveness to ozone in nonsmokers and smokers. Boston, MA: Health Effects Institute.
Frampton. MW: Pryor. WA; Cueto. R; Cox. C: Morrow. PE; Utell. MJ. (1999). Ozone exposure increases
aldehydes in epithelial lining fluid in human lung. Am J Respir Crit Care Med 159: 1134-1137.
Frank. R; Liu. MC: Spannhake. EW: Mlynarek. S: Macri. K; Weinmann. GG. (2001). Repetitive ozone
exposure of young adults: Evidence of persistent small airway dysfunction. Am J Respir Crit Care Med
164: 1253-1260.
5-86
-------
Freed. AN; Chou. CL; Fuller. SD; Croxton. TL. (1996). Ozone-induced vagal reflex modulates airways
reactivity in rabbits. Respir Physiol Neurobiol 105: 95-102.
Freed. AN; Cueto. R; Pryor. WA. (1999). Antioxidant transport modulates peripheral airway reactivity and
inflammation during ozone exposure. J Appl Physiol 87: 1595-1603.
Fujita. M; Sasavama. S: Ohno. A; Nakajima. H; Asanoi. H. (1987). Importance of angina for development of
collateral circulation. Heart 57: 139-143.
Gackiere. F; Saliba. L; Baude. A; Bosler. O; Strube. C. (2011). Ozone inhalation activates stress-responsive
regions of the CNS. JNeurochem 117: 961-972. http://dx.doi.org/10.1111/i.1471-4159.2011.07267.x
Gao. X; Raghavamenon. AC: D'Auvergne. O: Uppu. PJVI. (2009b). Cholesterol secoaldehyde induces
apoptosis in J774 macrophages via mitochondria! pathway but not involving reactive oxygen species as
mediators. Biochem Biophys Res Commun 389: 382-387. http://dx.doi.Org/10.1016/i.bbrc.2009.09.005
Garantziotis. S: Li. Z; Potts. EN; Kimata. K; Zhuo. L; Morgan. PL; Savani. RC: Noble. PW: Foster. WM;
Schwartz. DA; Hollingsworth. JW. (2009). Hyaluronan mediates ozone-induced airway
hyperresponsiveness in mice. J Biol Chem 284: 11309-11317. http://dx.doi.org/10.1074/ibc.M802400200
Garantziotis. S: Li. Z; Potts. EN; Lindsev. JY; Stober. VP; Polosukhin. W; Blackwell. TS: Schwartz. DA;
Foster. WM; Hollingsworth. JW. (2010). TLR4 is necessary for hyaluronan-mediated airway
hyperresponsiveness after ozone inhalation. Am J Respir Crit Care Med 181: 666-675.
http://dx.doi.org/10.1164/rccm.200903-0381OC
Garcia. GJ: Schroeter. JD; Segal. RA; Stanek. J: Foureman. GL; Kimbell. JS. (2009a). Dosimetry of nasal
uptake of water-soluble and reactive gases: A first study of interhuman variability. Inhal Toxicol 21: 607-
618. http://dx.doi.org/10.1080/08958370802320186
Gerritv. TR; Biscardi. F; Strong. A; Garlington. AR; Brown. JS: Bromberg. PA. (1995). Bronchoscopic
determination of ozone uptake in humans. J Appl Physiol 79: 852-860.
Gerritv. TR; McDonnell. WF; House. DE. (1994). The relationship between delivered ozone dose and
functional responses in humans. Toxicol Appl Pharmacol 124: 275-283.
Gerritv. TR; Weaver. RA; Berntsen. J: House. DE; O'Neil. JJ. (1988). Extrathoracic and intrathoracic removal
of O3 in tidal-breathing humans. J Appl Physiol 65: 393-400.
Giamalva. D; Church. DF; Pryor. WA. (1985). A comparison of the rates of ozonation of biological
antioxidants and oleate and linoleate esters. Biochem Biophys Res Commun 133: 773-779.
Gong. H. Jr; Wong. R; Sarma. RJ: Linn. WS: Sullivan. ED; Shamoo. DA; Anderson. KR; Prasad. SB. (1998).
Cardiovascular effects of ozone exposure in human volunteers. Am J Respir Crit Care Med 158: 538-546.
Graham. DE; Koren. HS. (1990). Biomarkers of inflammation in ozone-exposed humans: Comparison of the
nasal and bronchoalveolar lavage. Am J Respir Crit Care Med 142: 152-156.
Gross. EA; Starr. TB; Randall. HW: Morgan. KT. (1987). Morphometric analysis of the primate nasal cavity
[Abstract]. Toxicologist 7: 193.
Gross. EA; Swenberg. JA; Fields. S: Popp. JA. (1982). Comparative morphometry of the nasal cavity in rats
and mice. JAnat 135: 83-88.
Guevara-Guzman. R; Arriaga. V: Kendrick. KM; Bernal. C: Vega. X; Mercado-Gomez. OF; Rivas-Arancibia.
S.. (2009). Estradiol prevents ozone-induced increases in brain lipid peroxidation and impaired social
recognition memory in female rats. Neuroscience 159: 940-950.
http://dx.doi.0rg/10.1016/i.neuroscience.2009.01.047
Gunnison. AF; Hatch. GE. (1999). 03-induced inflammation in prepregnant, pregnant, and lactating rats
correlates with O3 dose estimated by 18O. Am J Physiol 276: L332-L340.
Gunnison. AF; Hatch. GE; Crissman. K; Bowers. A. (1996). Comparative sensitivity of lactating and virgin
female rats to ozone-induced pulmonary inflammation. Inhal Toxicol 8: 607-623.
5-87
-------
Hackney. JD; Linn. WS: Mohler. JG: Collier. CR. (1977b). Adaptation to short-term respiratory effects of
ozone in men exposed repeatedly. J Appl Physiol 43: 82-85.
Hamade. AK; Misra. V: Rabold. R; Tankerslev. CG. (2010). Age-related changes in cardiac and respiratory
adaptation to acute ozone and carbon black exposures: Interstrain variation in mice. Inhal Toxicol 22: 84-
94. http://dx.doi.org/10.3109/08958378.2010.503974
Hamade. AK; Tankerslev. CG. (2009). Interstrain variation in cardiac and respiratory adaptation to repeated
ozone and particulate matter exposures. Am J Physiol Regul Integr Comp Physiol 296: R1202-R1215.
http://dx.doi.org/10.1152/ajpregu.90808.2008
Hamilton. RF; Li. L; Eschenbacher. WL; Szweda. L; Holian. A. (1998). Potential involvement of 4-
hydroxynonenal in the response of human lung cells to ozone. Am J Physiol 274: L8-L16.
Han. SG: Andrews. R; Gairola. CG: Bhalla. DK. (2008). Acute pulmonary effects of combined exposure to
carbon nanotubes and ozone in mice. Inhal Toxicol 20: 391-398.
http://dx.doi.org/10.1080/08958370801904014
Hansen. CA; Barnett. AG: Pritchard. G. (2008). The effect of ambient air pollution during early pregnancy on
fetal ultrasonic measurements during mid-pregnancy. Environ Health Perspect 116: 362-369.
http://dx.doi.org/10.1289/ehp.10720
Hague. R: Umstead. TM; Freeman. WM; Floros. J: Phelps. PS. (2009). The impact of surfactant protein-A on
ozone-induced changes in the mouse bronchoalveolar lavage proteome. Proteome Science 7: 12.
http://dx.doi.org/10.1186/1477-5956-7-12
Hague. R: Umstead. TM; Ponnuru. P; Guo. X; Hawgood. S: Phelps. PS; Floros. J. (2007). Role of surfactant
protein-A (SP-A) in lung injury in response to acute ozone exposure of SP-A deficient mice. Toxicol Appl
Pharmacol 220: 72-82. http://dx.doi.Org/10.1016/i.taap.2006.12.017
Harkema. JR; Hotchkiss. JA; Barr. EB; Bennett. CB; Gallup. M; Lee. JK; Basbaum. C. (1999). Long-lasting
effects of chronic ozone exposure on rat nasal epithelium. Am J Respir Cell Mol Biol 20: 517-529.
Harkema. JR: Plopper. CG: Hyde. DM; St George. JA; Pungworth. PL. (1987a). Effects of an ambient level of
ozone on primate nasal epithelial mucosubstances: quantitative histochemistry. Am J Pathol 127: 90-96.
Harkema. JR: Plopper. CG: Hyde. DM; St George. JA: Wilson. DW: Dungworth. PL. (1987b). Response of
the macaque nasal epithelium to ambient levels of ozone: A morphologic and morphometric study of the
transitional and respiratory epithelium. Am J Pathol 128: 29-44.
Harkema. JR: Plopper. CG: Hyde. PM; St George. JA: Wilson. PW: Pungworth. PL. (1993). Response of
macaque bronchiolar epithelium to ambient concentrations of ozone. Am J Pathol 143: 857-866.
Harkema. JR: Wagner. JG. (2005). Epithelial and inflammatory responses in the airways of laboratory rats
coexposed to ozone and biogenic substances: Enhancement of toxicant-induced airway injury. Exp Toxicol
Pathol 57: 129-141. http://dx.doi.Org/10.1016/i.etp.2005.05.013
Hatch. GE. (1992). Comparative biochemistry of airway lining fluid. In RA Parent (Ed.), Comparative biology
of the normal lung, vol 1: Treatise on pulmonary toxicology (pp. 617-632). Boca Raton, FL: CRC Press,
Inc.
Hatch. GE; Slade. R: Harris. LP; Mcdonnell. WF; Pevlin. RB; Koren. HS; Costa. PL; Mckee. J. (1994).
Ozone dose and effect in humans and rats: A comparison using oxygen-18 labeling and bronchoalveolar
lavage. Am J Respir Crit Care Med 150: 676-683.
Hatch. GE; Slade. R: Stead. AG: Graham. JA. (1986). Species comparison of acute inhalation toxicity of ozone
and phosgene. J Toxicol Environ Health 19: 43-53. http://dx.doi.org/10.1080/15287398609530905
Hatch. GE; Wiester. MJ; Overton. JH. Jr; Aissa. M. (1989). Respiratory tract dosimetry of [18]O-labeled ozone
in rats: Implications for a rat-human extrapolation of ozone dose. In T Schneider; SP Lee; GJR Wolters;
LP Grant (Eds.), Atmospheric ozone research and its policy implications (pp. 553-560). Amsterdam,
The Netherlands: Elsevier Science Publishers B.V.
http://www.elsevier.com/wps/find/bookdescription.cws home/502592/description#description
-------
Hazbun. ME; Hamilton. R; Holian. A; Eschenbacher. WL. (1993). Ozone-induced increases in substance P and
8-epi-prostaglandin F2 alpha in the airways of human subjects. Am J Respir Cell Mol Biol 9: 568-572.
http://dx.doi.Org/10.1165/ajrcmb/9.5.568
Hazucha. MJ; Bates. DV: Bromberg. PA. (1989). Mechanism of action of ozone on the human lung. J Appl
Physiol67: 1535-1541.
Hazucha. MJ: Folinsbee. LJ; Bromberg. PA. (2003). Distribution and reproducibility of spirometric response to
ozone by gender and age. J Appl Physiol 95: 1917-1925.
Hazucha. MJ: Madden. M; Pape. G: Becker. S: Devlin. R; Koren. HS: Kehrl. H; Bromberg. PA. (1996). Effects
of cyclo-oxygenase inhibition on ozone-induced respiratory inflammation and lung function changes. Eur J
Appl Physiol 73: 17-27.
Hernandez. ML; Lav. JC: Harris. B; Esther. CR; Brickev. WJ: Bromberg. PA; Diaz-Sanchez. D; Devlin. RB;
Kleeberger. SR; Alexis. NE; Peden. DB. (2010). Atopic asthmatic subjects but not atopic subjects without
asthma have enhanced inflammatory response to ozone. J Allergy Clin Immunol 126: 537-544.
http://dx.doi.0rg/10.1016/i.iaci.2010.06.043
Hoigne. J: Bader. H. (1983). Rate constants of reactions of ozone with organic and inorganic compounds in
water - II: Dissociating organic compounds. Water Res 17: 185-194. http://dx.doi.org/10.1016/0043-
1354(83)90099-4
Hollingsworth. JW: Cook. DN; Brass. DM; Walker. JKL; Morgan. PL; Foster. WM; Schwartz. DA. (2004).
The role of Toll-like receptor 4 in environmental airway injury in mice. Am J Respir Crit Care Med 170:
126-132. http://dx.doi.org/10.1164/rccm.200311-1499OC
Hollingsworth. JW: Free. ME; Li. Z; Andrews. LN; Nakano. H; Cook. DN. (2010). Ozone activates pulmonary
dendritic cells and promotes allergic sensitization through a Toll-like receptor 4-dependent mechanism
[Letter]. J Allergy Clin Immunol 125: 1167-1170. http://dx.doi.Org/10.1016/i.iaci.2010.03.001
Hollingsworth. JW: Maruoka. S: Li. Z; Potts. EN; Brass. DM; Garantziotis. S: Fong. A; Foster. WM;
Schwartz. DA. (2007). Ambient ozone primes pulmonary innate immunity in mice. J Immunol 179: 4367-
4375.
Holtzman. MJ: Cunningham. JH; Sheller. JR; Irsigler. GB; Nadel. JA; Boushev. HA. (1979). Effect of ozone
on bronchial reactivity in atopic and nonatopic subjects. Am Rev Respir Dis 120: 1059-1067.
Holtzman. MJ: Fabbri. LM; O'Byrne. PM; Gold. BD; Aizawa. H; Walters. EH; Alpert. SE; Nadel. JA. (1983).
Importance of airway inflammation for hyperresponsiveness induced by ozone. Am Rev Respir Dis 127:
686-690.
Holz. O; Jorres. RA; Timm. P; Mucke. M; Richter. K; Koschyk. S: Magnussen. H. (1999). Ozone-induced
airway inflammatory changes differ between individuals and are reproducible. Am J Respir Crit Care Med
159: 776-784.
Horstman. DH; Ball. BA; Brown. J: Gerritv. T; Folinsbee. LJ. (1995). Comparison of pulmonary responses of
asthmatic and nonasthmatic subjects performing light exercise while exposed to a low level of ozone.
Toxicol Ind Health 11: 369-385.
Horvath. SM; Gliner. JA; Folinsbee. LJ. (1981). Adaptation to ozone: Duration of effect. Am Rev Respir Dis
123: 496-499.
Hotchkiss. JA; Harkema. JR: Henderson. RF. (1991). Effect of cumulative ozone exposure on ozone-induced
nasal epithelial hyperplasia and secretory metaplasia in rats. Exp Lung Res 15: 589-600.
Hu. PC: Miller. FJ; Daniels. MJ: Hatch. G. (1982). Protein accumulation in lung lavage fluid following ozone
exposure. Environ Res 29: 377-388. http://dx.doi.org/10.1016/0013-9351(82)90039-1
Hu. SC: Ben-Jebria. A; Ultman. JS. (1992). Longitudinal distribution of ozone absorption in the lung: Quiet
respiration in healthy subjects. J Appl Physiol 73: 1655-1667.
Hu. SC: Ben-Jebria. A; Ultman. JS. (1994). Longitudinal distribution of ozone absorption in the lung: Effects
of respiratory flow. J Appl Physiol 77: 574-583.
5-89
-------
Ichinose. T; Arakawa. K; Shimojo. N; Sagai. M. (1988). Biochemical effects of combined gases of nitrogen
dioxide and ozone: II Species differences in lipid peroxides and antioxidative protective enzymes in the
lungs. Toxicol Lett 42: 167-176.
ICRP (International Commission on Radiological Protection). (1994). Human respiratory tract model for
radiological protection: A report of a task group of the International Commission on Radiological
Protection. ICRP Publication 66 [Review]. Ann ICRP 24: 1-482.
Ignatenko. AV; Cherenkevich. SN. (1985). Reactivity of amino-acids and proteins in reactions with ozone.
Kinet Catal 26: 1145-1148.
Islam. T; Mcconnell. R; Gauderman. WJ: Avol. E; Peters. JM; Gilliland. FD. (2008). Ozone, oxidant defense
genes and risk of asthma during adolescence. Am J Respir Crit Care Med 177: 388-395.
http://dx.doi.org/10.1164/rccm.200706-863OC
Iwasaki. T; Takahashi. M; Saito. H; Arito. H. (1998). Adaptation of extrapulmonary responses to ozone
exposure in conscious rats. Ind Health 36: 57-60.
Jalaludin. B; Mamies. T; Morgan. G: Lincoln. D; Sheppeard. V: Corbett. S. (2007). Impact of ambient air
pollution on gestational age is modified by season in Sydney, Australia. Environ Health 6: 16.
http://dx.doi.org/10.1186/1476-069X-6-16
James. PS: Stidlev. CA; Lambert. WE; Chick. TW: Mermier. CM; Samet. JM. (1997). Oronasal distribution of
ventilation at different ages. Arch Environ Occup Health 52: 118-123.
Jiang. D; Liang. J; Fan. J; Yu. S: Chen. S: Luo. Y; Prestwich. GD; Mascarenhas. MM; Garg. HG: Quinn. DA;
Homer. RJ; Goldstein. PR; Bucala. R; Lee. PJ: Medzhitov. R; Noble. PW. (2005). Regulation of lung injury
and repair by Toll-like receptors and hyaluronan. Nat Med 11: 1173-1179.
http://dx.doi.org/10.1038/nml315
Joad. JP; Kott. KS: Brie. JM. (1996). The local C-fiber contribution to ozone-induced effects on the isolated
guinea pig lung. Toxicol Appl Pharmacol 141: 561-567.
Joad. JP: Kott. KS: Brie. JM; Peake. JL; Plopper. CG: Schelegle. ES: Gershwin. LJ; Pinkerton. KE. (2006).
Structural and functional localization of airway effects from episodic exposure of infant monkeys to
allergen and/or ozone. Toxicol Appl Pharmacol 214: 237-243. http://dx.doi.Org/10.1016/i.taap.2005.12.012
Johansson. E; Wesselkamper. SC: Shertzer. HG: Leikauf. GD: Dalton. TP; Chen. Y. (2010). Glutathione
deficient C57BL/6J mice are not sensitized to ozone-induced lung injury. Biochem Biophys Res Commun
396: 407-412. http://dx.doi.Org/10.1016/i.bbrc.2010.04.105
Johnston. RA; Mizgerd. JP: Shore. SA. (2005 a). CXCR2 is essential for maximal neutrophil recruitment and
methacholine responsiveness after ozone exposure. Am J Physiol Lung Cell Mol Physiol 288: L61-L67.
http://dx.doi.org/10.1152/aiplung.00101.2004 00101.2004
Johnston. RA; Schwartzman. IN; Flynt. L; Shore. SA. (2005b). Role of interleukin-6 in murine airway
responses to ozone. Am J Physiol Lung Cell Mol Physiol 288: L390-L397.
http://dx.doi.org/10.1152/aiplung.00007.2004
Johnston. RA; Theman. TA; Lu. FL; Terry. RD; Williams. ES: Shore. SA. (2008). Diet-induced obesity causes
innate airway hyperresponsiveness to methacholine and enhances ozone-induced pulmonary inflammation.
J Appl Psychol 104: 1727-1735. http://dx.doi.org/10.1152/iapplphysiol.00075.2008
Jorres. R: Nowak. D; Magnussen. H; Speckin. P; Koschyk. S. (1996). The effect of ozone exposure on allergen
responsiveness in subjects with asthma or rhinitis. Am J Respir Crit Care Med 153: 56-64.
Jorres. RA; Holz. O; Zachgo. W: Timm. P; Koschyk. S: Muller. B; Grimminger. F; Seeger. W: Kelly. FJ;
Punster. C: Frischer. T; Lubec. G: Waschewski. M; Niendorf. A; Magnussen. H. (2000). The effect of
repeated ozone exposures on inflammatory markers in bronchoalveolar lavage fluid and mucosal biopsies.
Am J Respir Crit Care Med 161: 1855-1861.
Kabel. JR; Ben-Jebria. A; Ultman. JS. (1994). Longitudinal distribution of ozone absorption in the lung:
Comparison of nasal and oral quiet breathing. J Appl Physiol 77: 2584-2592.
5-90
-------
Kafoury. RM; Pryor. WA; Squadrito. GL; Salgo. MG: Zou. X; Friedman. M. (1998). Lipid ozonation products
activate phospholipases A2, C, and D. Toxicol Appl Pharmacol 150: 338-349.
Kaiekar. R; Pieczarka. EM; Smiley-Jewell. SM; Schelegle. ES: Fanucchi. MV; Plopper. CG. (2007). Early
postnatal exposure to allergen and ozone leads to hyperinnervation of the pulmonary epithelium. Respir
Physiol Neurobiol 155: 55-63. http://dx.doi.Org/10.1016/i.resp.2006.03.002
Kannan. S: Misra. DP; Dvonch. T; Krishnakumar. A. (2006). Exposures to airborne paniculate matter and
adverse perinatal outcomes: A biologically plausible mechanistic framework for exploring potential effect
modification by nutrition. Environ Health Perspect 114: 1636-1642.
Kanofsky. JR; Sima. PP. (1995). Reactive absorption of ozone by aqueous biomolecule solutions: Implications
for the role of sulfhydryl compounds as targets for ozone. Arch Biochem Biophys 316: 52-62.
Kari. F; Hatch. G: Slade. R; Crissman. K; Simeonova. PP; Luster. M. (1997). Dietary restriction mitigates
ozone-induced lung inflammation in rats: A role for endogenous antioxidants. Am J Respir Cell Mol Biol
17: 740-747.
Katre. A; Ballinger. C: Akhter. H; Fanucchi. M; Kim. DK; Postlethwait E; Liu. RM. (2011). Increased
transforming growth factor beta 1 expression mediates ozone-induced airway fibrosis in mice. Inhal
Toxicol 23: 486-494. http://dx.doi.org/10.3109/08958378.2011.584919
Kavlock. R; Daston. G: Grabowski. CT. (1979). Studies on the developmental toxicity of ozone. I. Prenatal
effects. Toxicol Appl Pharmacol 48: 19-28. http://dx.doi.org/10.1016/S0041-008X(79)80004-6
Kehrl. HR; Peden. DB; Ball. BA; Folinsbee. LJ; Horstman. DH. (1999). Increased specific airway reactivity of
persons with mild allergic asthma after 7.6 hours of exposure to 0.16 ppm ozone. J Allergy Clin Immunol
104: 1198-1204.
Kehrl. HR; Vincent. LM; Kowalsky. RJ; Horstman. DH; O'Neil. JJ: McCartney. WH; Bromberg. PA. (1987).
Ozone exposure increases respiratory epithelial permeability in humans. Am Rev Respir Dis 135: 1124-
1128.
Kermani. S: Ben-Jebria. A; Ultman. JS. (2006). Kinetics of ozone reaction with uric acid, ascorbic acid, and
glutathione at physiologically relevant conditions. Arch Biochem Biophys 451: 8-16.
htrp://dx.doi.org/10.1016/i.abb.2006.04.015
Kim. CS: Alexis. NE; Rappold. AG: Kehrl. H; Hazucha. MJ; Lav. JC: Schmitt. MT; Case. M; Devlin. RB;
Peden. DB; Diaz-Sanchez. D. (2011). Lung function and inflammatory responses in healthy young adults
exposed to 0.06 ppm ozone for 6.6 hours. Am J Respir Crit Care Med 183: 1215-1221.
http://dx.doi.org/10.1164/rccm.201011-1813OC
Kleeberger. SR; Levitt. RC: Zhang. LY; Longphre. M; Harkema. J: Jedlicka. A; Eleff. SM; DiSilvestre. D:
Holroyd. KJ. (1997). Linkage analysis of susceptibility to ozone-induced lung inflammation in inbred mice.
Nat Genet 17: 475-478.
Kleeberger. SR: Reddv. S: Zhang. LY; Jedlicka. AE. (2000). Genetic susceptibility to ozone-induced lung
hyperpermeability: Role of toll-like receptor 4. Am J Respir Cell Mol Biol 22: 620-627.
Kleeberger. SR: Reddv. SP; Zhang. LY; Cho. HY; Jedlicka. AE. (2001). Toll-like receptor 4 mediates ozone-
induced murine lung hyperpermeability via inducible nitric oxide synthase. Am J Physiol 280: L326-L333.
Kleeberger. SR: Seiden. JE; Levitt. RC: Zhang. LY. (1993). Mast cells modulate acute ozone-induced
inflammation of the murine lung. Am J Respir Crit Care Med 148: 1284-1291.
Kodavanti. UP; Costa. PL; Dreher. KL; Crissman. K; Hatch. GE. (1995). Ozone-induced tissue injury and
changes in antioxidant homeostasis in normal and ascorbate-deficient guinea pigs. Biochem Pharmacol 50:
243-251. htrp://dx.doi.org/10.1016/0006-2952(95)00122-G
Kodavanti. UP; Thomas. R: Ledbetter. AD; Schladweiler. MC: Shannahan. JH; Wallenborn. JG: Lund. AK;
Campen. MJ: Butler. EO; Gottipolu. RR; Nyska. A; Richards. JE; Andrews. D: Jaskot. RH; Mckee. J:
Kotha. S. R.: Patel. RB; Parianandi. NL. (2011). Vascular and cardiac impairments in rats Inhaling ozone
and diesel exhaust particles. Environ Health Perspect 119: 312-318. http://dx.doi.org/10.1289/ehp.1002386
5-91
-------
Koenig. JQ; Covert. DS: Marshall. SG: Van Belle. G: Pierson. WE. (1987). The effects of ozone and nitrogen
dioxide on pulmonary function in healthy and in asthmatic adolescents. Am J Respir Crit Care Med 136:
1152-1157.
Koren. HS: Devlin. RB: Graham. DE; Mann. R; Mcgee. MP; Horstman. DH; Kozumbo. WJ: Becker. S:
House. DE; Mcdonnell. WF; Bromberg. PA. (1989). Ozone-induced inflammation in the lower airways of
human subjects. Am J Respir Crit Care Med 139: 407-415.
Kreit. JW: Gross. KB; Moore. TB; Lorenzen. TJ; D'Arcv. J: Eschenbacher. WL. (1989). Ozone-induced
changes in pulmonary function and bronchial responsiveness in asthmatics. J Appl Physiol 66: 217-222.
Krishna. MT; Springall. D; Meng. QH; Withers. N; Macleod. D; Biscione. G: Frew. A; Polak. J: Holgate. S.
(1997). Effects of ozone on epithelium and sensory nerves in the bronchial mucosa of healthy humans. Am
J Respir Crit Care Med 156: 943-950.
Larsen. ST; Matsubara. S: Mcconville. G: Poulsen. SS: Gelfand. EW. (2010). Ozone increases airway
hyperreactivity and mucus hyperproduction in mice previously exposed to allergen. J Toxicol Environ
Health A 73: 738-747. http://dx.doi.org/10.1080/15287391003614034
Larson. SD; Schelegle. ES: Walbv. WF; Gershwin. LJ; Fanuccihi. MV; Evans. MJ; Joad. JP; Tarkington. BK;
Hyde. DM; Plopper. CG. (2004). Postnatal remodeling of the neural components of the epithelial-
mesenchymal trophic unit in the proximal airways of infant rhesus monkeys exposed to ozone and allergen.
Toxicol Appl Pharmacol 194: 211-220. http://dx.doi.Org/10.1016/j.taap.2003.09.025
Lav. JC: Alexis. NE; Kleeberger. SR; Roubev. RA; Harris. BD; Bromberg. PA; Hazucha. MJ: Devlin. RB;
Peden. DB. (2007). Ozone enhances markers of innate immunity and antigen presentation on airway
monocytes in healthy individuals [Letter]. J Allergy Clin Immunol 120: 719-722.
http://dx.doi.0rg/10.1016/i.jaci.2007.05.005
Li. YF; Gauderman. WJ: Avol. E; Dubeau. L; Gilliland. FD. (2006d). Associations of tumor necrosis factor G-
308A with childhood asthma and wheezing. Am J Respir Crit Care Med 173: 970-976.
http://dx.doi.org/10.1164/rccm.200508-1256OC
Li. Z; Potts-Kant. EN; Garantziotis. S: Foster. WM; Hollingsworth. JW. (2011). Hyaluronan signaling during
ozone-induced lung injury requires TLR4, MyD88, and TIRAP PLoS ONE 6: e27137.
http://dx.doi.org/10.1371/iournal.pone.0027137
Li. Z; Potts. EN; Piantadosi. CA; Foster. WM; Hollingsworth. JW. (2010). Hyaluronan fragments contribute to
the ozone-primed immune response to lipopolysaccharide. J Immunol 185: 6891-6898.
http://dx.doi.org/10.4049/iimmunol.1000283
Linn. WS: Buckley. RD; Spier. CE; Blessev. RL; Jones. MP; Fischer. DA; Hackney. JD. (1978). Health effects
of ozone exposure in asthmatics. Am Rev Respir Dis 117: 835-843.
Linn. WS: Medwav. DA; Anzar. UT; Valencia. LM; Spier. CE; FS-D. T; Fischer. DA; Hackney. JD. (1982b).
Persistence of adaptation to ozone in volunteers exposed repeatedly for six weeks. Am Rev Respir Dis 125:
491-495.
Liu. L; Leech. JA; Urch. RB: Poon. R; Zimmerman. B; Kubav. JM; Silverman. FS. (1999). A comparison of
biomarkers of ozone exposure in human plasma, nasal lavage, and sputum. Inhal Toxicol 11: 657-674.
http://dx.doi.org/10.1080/089583799196790
Liu. L; Leech. JA; Urch. RB: Silverman. FS. (1997). In vivo salicylate hyroxylation: A potential biomarker for
assessing acute ozone exposure and effects in humans. Am J Respir Crit Care Med 156: 1405-1412.
London. SJ. (2007). Gene-air pollution interactions in asthma [Review]. Proc Am Thorac Soc 4: 217-220.
http://dx.doi.org/10.1513/pats.200701-031AW
Long. NC: Suh. J: Morrow. JD: Schiestl. RH; Krishna Murthy. GG: Brain. JD: Frei. B. (2001). Ozone causes
lipid peroxidation but little antioxidant depletion in exercising and nonexercising hamsters. J Appl Physiol
91: 1694-1700.
Longphre. M; Zhang. LY; Harkema. JR; Kleeberger. SR. (1999). Ozone-induced pulmonary inflammation and
epithelial proliferation are partially mediated by PAF. J Appl Physiol 86: 341-349.
5-92
-------
Maniar-Hew. K; Postlethwait EM; Fanucchi. MV; Ballinger. CA; Evans. MJ; Harkema. JR: Carey. SA;
McDonald. RJ: Bartolucci. AA; Miller. LA. (2011). Postnatal episodic ozone results in persistent
attenuation of pulmonary and peripheral blood responses to LPS challenge. Am J Physiol Lung Cell Mol
Physiol 300: L462-L471. http://dx.doi.org/10.1152/ajplung.00254.2010
Matsumura. Y. (1970). The effects of ozone, nitrogen dioxide, and sulfur dioxide on the experimentally
induced allergic respiratory disorder in guinea pigs: II. The effects of ozone on the absorption and the
retention of antigen in the lung. Am Rev Respir Dis 102: 438-443.
McBride. JT. (1992). Architecture of the tracheobronchial tree. In RA Parent (Ed.), Comparative biology of the
normal lung, v 1: Treatise on pulmonary toxicology (pp. 49-61). Boca Raton, FL: CRC Press.
McDonnell. WF; Horstman. DH; Hazucha. MJ: Seal. E. Jr; Haak. ED; Salaam. SA; House. DE. (1983).
Pulmonary effects of ozone exposure during exercise: Dose-response characteristics. J Appl Physiol 54:
1345-1352.
Mckinnev. WJ: Jaskot. RH; Richards. JH; Costa. PL; Dreher. KL. (1998). Cytokine mediation of ozone-
induced pulmonary adaptation. Am J Respir Cell Mol Biol 18: 696-705.
Mercer. RR; Anjilvel. S: Miller. FJ; Crapo. JD. (1991). Inhomogeneity of ventilatory unit volume and its
effects on reactive gas uptake. J Appl Physiol 70: 2193-2205.
Mercer. RR; Russell. ML; Crapo. JD. (1992). Mucous lining layers in human and rat airways [Abstract]. Am
Rev Respir Dis 145: A355.
Mikerov. AN; Gan. X; Umstead. TM; Miller. L; Chinchilli. VM; Phelps. PS: Floros. J. (2008a). Sex
differences in the impact of ozone on survival and alveolar macrophage function of mice after Klebsiella
pneumoniae infection. Respir Res 9: 24. http://dx.doi.org/10.1186/1465-9921-9-24
Mikerov. AN; Hague. R; Gan. X; Quo. X; Phelps. PS: Floros. J. (2008b). Ablation of SP-A has a negative
impact on the susceptibility of mice to Klebsiella pneumoniae infection after ozone exposure: Sex
differences. Respir Res 9: 77. http://dx.doi.org/10.1186/1465-9921-9-77
Mikerov. AN; Umstead. TM; Gan. X; Huang. W: Quo. X; Wang. G: Phelps. PS: Floros. J. (2008c). Impact of
ozone exposure on the phagocytic activity of human surfactant protein A (SP-A) and SP-A variants. Am J
Physiol Lung Cell Mol Physiol 294: L121-L130. http://dx.doi.org/10.1152/aiplung.00288.2007
Miller. FJ. (1995). Uptake and fate of ozone in the respiratory tract [Review]. Toxicol Lett 82-83: 277-285.
Miller. FJ: Kimbell. JS. (1995). Regional dosimetry of inhaled reactive gases. In RO McClellan; RF Henderson
(Eds.), Concepts in inhalation toxicology (2nd ed., pp. 257-287). Washington, DC: Taylor & Francis.
Miller. FJ: McNeal. CA; Kirtz. JM; Gardner. DE; Coffin. PL; Menzel. DB. (1979). Nasopharyngeal removal
of ozone in rabbits and guinea pigs. Toxicology 14: 273-281. http://dx.doi.org/10.1016/0300-
483X(79)90009-X
Miller. FJ: Overton. JH; Gerritv. TR; Graham. RC. (1988). Interspecies dosimetry of reactive gases. In U
Mohr; D Dungworth; R McClellan; G Kimmerle; W Stober; J Lewkowski (Eds.), Inhalation toxicology:
The design and interpretation of inhalation studies and their use in risk assessment (pp. 139-155). New
York, NY: Springer-Verlag.
Miller. FJ: Overton. JH. Jr; Jaskot. RH; Menzel. DB. (1985). Amodel of the regional uptake of gaseous
pollutants in the lung: I. The sensitivity of the uptake of ozone in the human lung to lower respiratory tract
secretions and exercise. Toxicol Appl Pharmacol 79: 11-27. http://dx.doi.org/10.1016/0041-
008X(85)90364-3
Miller. LA; Gerriets. JE; Tyler. NK; Abel. K; Schelegle. ES: Plopper. CG: Hyde. DM. (2009). Ozone and
allergen exposure during postnatal development alters the frequency and airway distribution of CD25+
cells in infant rhesus monkeys. Toxicol Appl Pharmacol 236: 39-48.
http://dx.doi.0rg/10.1016/i.taap.2008.12.031
Molfino. NA; Wright. SC: Katz. I; Tarlo. S: Silverman. F; Mcclean. PA; Szalai. JP; Raizenne. M; Slutsky. AS:
Zamel. N. (1991). Effect of low concentrations of ozone on inhaled allergen responses in asthmatic
subjects. Lancet 338: 199-203.
5-93
-------
Morgan. KT; Monticello. TM. (1990). Airflow, gas deposition, and lesion distribution in the nasal passages.
Environ Health Perspect 85: 209-218.
Mudwav. IS: Behndig. AF; Helledav. R; Pourazar. J; Frew. AJ; Kelly. FJ; Blomberg. A. (2006). Vitamin
supplementation does not protect against symptoms in ozone-responsive subjects. Free Radic Biol Med 40:
1702-1712.
Mudwav. IS: Blomberg. A; Frew. AJ: Holgate. ST; Sandstrom. T; Kelly. FJ. (1999a). Antioxidant consumption
and repletion kinetics in nasal lavage fluid following exposure of healthy human volunteers to ozone. Eur
Respir J 13: 1429-1438.
Mudwav. IS: Houslev. D; Eccles. R; Richards. RJ: Datta. AK; Tetlev. TD; Kelly. FJ. (1996). Differential
depletion of human respiratory tract antioxidants in response to ozone challenge. Free Radic Res 25: 499-
513.
Mudwav. IS: Kelly. FJ. (1998). Modeling the interactions of ozone with pulmonary epithelial lining fluid
antioxidants. Toxicol Appl Pharmacol 148: 91-100.
Mudwav. IS: Kelly. FJ. (2000). Ozone and the lung: A sensitive issue. Mol Aspects Med 21: 1-48.
Mudwav. IS: Kelly. FJ. (2004b). An investigation of inhaled ozone dose and the magnitude of airway
inflammation in healthy adults: Online data supplement. Am J Respir Crit Care Med 169: 1089-1095.
http://dx.doi.org/10.1164/rccm.200309-1325PP
Mudwav. IS: Krishna. MT; Frew. AJ: MacLeod. D; Sandstrom. T; Holgate. ST; Kelly. FJ. (1999b).
Compromised concentrations of ascorbate in fluid lining the respiratory tract in human subjects after
exposure to ozone. Occup Environ Med 56: 473-481.
Mudwav. IS: Stenfors. N; Blomberg. A; Helledav. R: Punster. C: Marklund. SL; Frew. AJ: Sandstrom. T;
Kelly. FJ. (2001). Differences in basal airway antioxidant concentrations are not predictive of individual
responsiveness to ozone: A comparison of healthy and mild asthmatic subjects. Free Radic Biol Med 31:
962-974.
Murphy. RC: Johnson. KM. (2008). Cholesterol, reactive oxygen species, and the formation of biologically
active mediators. J Biol Chem 283: 15521-15525. http://dx.doi.org/10.1074/ibc.R700049200
Myatt L; Kossenjans. W: Sahav. R: Eis. A; Brockman. D. (2000). Oxidative stress causes vascular dysfunction
in the placenta [Review]. J Matern Fetal Med 9: 79-82. http://dx.doi.org/10.1002/(SICI)1520-
6661 (200001/02)9: K79:: AID-MFM16>3.0. CO;2-O
Nishiyama. H; Ikeda. H; Kaneko. T; Fu. L; Kudo. M; Ito. T; Okubo. T. (1998). Neuropeptides mediate the
ozone-induced increase in the permeability of the tracheal mucosa in guinea pigs. Am J Physiol 275: L231-
L238.
Nodelman. V; Ultman. JS. (1999). Longitudinal distribution of chlorine absorption in human airways:
A comparison to ozone absorption. J Appl Physiol 87: 2073-2080.
Noviski. N; Brewer. JP; Skornik. WA; Galli. SJ: Drazen. JM; Martin. TR. (1999). Mast cell activation is not
required for induction of airway hyperresponsiveness by ozone in mice. J Appl Physiol 86: 202-210.
O'Byrne. P; Walters. E; Gold. B; Aizawa. H; Fabbri. L; Alpert. S: Nadel. J: Holtzman. M. (1983). Neutrophil
depletion inhibits airway hyperresponsiveness induced by ozone exposure. Am Rev Respir Dis 130: 214-
219.
O'Byrne. PM; Walters. EH; Aizawa. H; Fabbri. LM; Holtzman. MJ; Nadel. JA. (1984). Indomethacin inhibits
the airway hyperresponsiveness but not the neutrophil influx induced by ozone in dogs. Am Rev Respir Dis
130: 220-224.
Oslund. KL; Hyde. DM; Putney. LF; Alfaro. MF; Walbv. WF; Tyler. NK; Schelegle. ES. (2008). Activation of
neurokinin-1 receptors during ozone inhalation contributes to epithelial injury and repair. Am J Respir Cell
Mol Biol 39: 279-288. http://dx.doi.org/10.1165/rcmb.2008-0009OC
5-94
-------
Oslund. KL; Hyde. DM; Putney. LF; Alfaro. MF; Walbv. WF; Tyler. NK; Schelegle. ES. (2009). Activation of
calcitonin gene-related peptide receptor during ozone inhalation contributes to airway epithelial injury and
repair. Toxicol Pathol 37: 805-813. http://dx.doi.org/10.1177/0192623309345691
Overton. JH; Graham. RC. (1989). Predictions of ozone absorption in human lungs from newborn to adult.
Health Phys 1:29-36.
Overton. JH; Graham. RC: Menache. MG: Mercer. RR; Miller. FJ. (1996). Influence of tracheobronchial
region expansion and volume on reactive gas uptake and interspecies dose extrapolations. Inhal Toxicol 8:
723-745.
Overton. JH: Graham. RC: Miller. FJ. (1987). Amodel of the regional uptake of gaseous pollutants in the lung:
II. The sensitivity of ozone uptake in laboratory animal lungs to anatomical and ventilatory parameters.
Toxicol Appl Pharmacol 88: 418-432. htrp://dx.doi.org/10.1016/0041-008X(87)90216-X
Paquette. NC: Zhang. LY; Ellis. WA; Scott. AL; Kleeberger. SR. (1996). Vitamin A deficiency enhances ozone-
induced lung injury. Am J Physiol 270: L475-L482.
Park. JW: Taube. C: Swasev. C: Kodama. T; Joetham. A; Balhorn. A; Takeda. K; Miyahara. N; Allen. CB;
Dakhama. A; Kim. SH; Dinarello. CA; Gelfand. EW. (2004). Interleukin-1 receptor antagonist attenuates
airway hyperresponsiveness following exposure to ozone. Am J Respir Cell Mol Biol 30: 830-836.
http://dx.doi.org/10.1165/rcmb.2003-0373OC
Passannante. AN; Hazucha. MJ; Bromberg. PA; Seal. E; Folinsbee. L; Koch. G. (1998). Nociceptive
mechanisms modulate ozone-induced human lung function decrements. J Appl Physiol 85: 1863-1870.
Pearson. AC: Bhalla. DK. (1997). Effects of ozone on macrophage adhesion in vitro and epithelial and
inflammatory responses in vivo: The role of cytokines. J Toxicol Environ Health 50: 143-157.
Peden. DB. (2011). The role of oxidative stress and innate immunity in O3 and endotoxin-induced human
allergic airway disease [Review]. Immunol Rev 242: 91-105. http://dx.doi.org/10.1111/i.1600-
065X.2011.01035.x
Peden. DB; Boehlecke. B; Horstman. D; Devlin. R. (1997). Prolonged acute exposure to 0.16 ppm ozone
induces eosinophilic airway inflammation in asthmatic subjects with allergies. J Allergy Clin Immunol 100:
802-808.
Peden. DB: Setzer. RW. Jr; Devlin. RB. (1995). Ozone exposure has both a priming effect on allergen-induced
responses and an intrinsic inflammatory action in the nasal airways of perennially allergic asthmatics. Am J
Respir Crit Care Med 151: 1336-1345.
Peden. DB: Swiersz. M; Ohkubo. K; Halm. B; Emery. B; Kaliner. MA. (1993). Nasal secretion of the ozone
scavenger uric acid. Am Rev Respir Dis 148: 455-461.
Perepu. RS: Garcia. C: Postal. D; Sethi. R. (2010). Enhanced death signaling in ozone-exposed ischemic-
reperfused hearts. Mol Cell Biochem 336: 55-64. http://dx.doi.org/10.1007/sll010-009-0265-4
Perevra-Munoz. N; Rugerio-Vargas. C: Angoa-Perez. M; Borgonio-Perez. G: Rivas-Arancibia. S. (2006).
Oxidative damage in substantia nigra and striatum of rats chronically exposed to ozone. J Chem Neuroanat
31: 114-123. http://dx.doi.0rg/10.1016/i.jchemneu.2005.09.006
Perez-Gil. J. (2008). Structure of pulmonary surfactant membranes and films: The role of proteins and lipid-
protein interactions [Review]. Biochim Biophys Acta 1778: 1676-1695.
http://dx.doi.0rg/10.1016/i.bbamem.2008.05.003
Pichavant. M; Goya. S: Meyer. EH; Johnston. RA; Kim. HY; Matangkasombut. P; Zhu. M; Iwakura. Y;
Savage. PB; Dekruyff. RH; Shore. SA; Umetsu. DT. (2008). Ozone exposure in a mouse model induces
airway hyperreactivity that requires the presence of natural killer T cells and IL-17. J Exp Med 205: 385-
393. http://dx.doi.org/10.1084/jem.20071507
Picher. M; Burch. LH; Boucher. RC. (2004). Metabolism of P2 receptor agonists in human airways:
Implications for mucociliary clearance and cystic fibrosis. J Biol Chem 279: 20234-20241.
http://dx.doi.org/10.1074/ibc.M400305200
5-95
-------
Plopper. CG: Hatch. GE; Wong. V; Duan. X; Weir. AJ; Tarkington. BK; Devlin. RB; Becker. S: Buckpitt. AR.
(1998). Relationship of inhaled ozone concentration to acute tracheobronchial epithelial injury, site-specific
ozone dose and glutathione depletion in rhesus monkeys. Am J Respir Cell Mol Biol 19: 387-399.
Plopper. CG: Mango. GW: Hatch. GE; Wong. VJ; Toskala. E; Reynolds. SD: Tarkington. BK; Stripp. BR.
(2006). Elevation of susceptibility to ozone-induced acute tracheobronchial injury in transgenic mice
deficient in Clara cell secretory protein. Toxicol Appl Pharmacol 213: 74-85.
http://dx.doi.0rg/10.1016/i.taap.2005.09.003
Plopper. CG: Smiley-Jewell. SM; Miller. LA; Fanucchi. MV; Evans. MJ; Buckpitt. AR; Avdalovic. M;
Gershwin. LJ; Joad. JP; Kaiekar. R; Larson. S: Pinkerton. KE; Van Winkle. LS: Schelegle. ES: Pieczarka.
EM; Wu. R; Hyde. DM. (2007). Asthma/allergic airways disease: Does postnatal exposure to
environmental toxicants promote airway pathobiology? Toxicol Pathol 35: 97-110.
http://dx.doi.org/10.1080/01926230601132030
Postlethwait. EM; Cueto. R: Velsor. LW: Pryor. WA. (1998). 03-induced formation of bioactive lipids:
Estimated surface concentrations and lining layer effects. Am J Physiol 274: L1006-L1016.
Postlethwait. EM: Joad. JP: Hyde. DM: Schelegle. ES: Brie. JM: Weir. AJ: Putney. LF: Wong. VJ: Velsor. LW:
Plopper. CG. (2000). Three-dimensional mapping of ozone-induced acute cytotoxicity in tracheobronchial
airways of isolated perfused rat lung. Am J Respir Cell Mol Biol 22: 191-199.
Postlethwait. EM; Langford. SD: Bidani. A. (1994). Determinants of inhaled ozone absorption in isolated rat
lungs. Toxicol Appl Pharmacol 125: 77-89.
Postlethwait. EM; Ultman. JS. (2001). Airspace surface chemistry mediates 03-induced lung injury. Hum Ecol
Risk Assess 7: 1145-1159. http://dx.doi.org/10.1080/20018091094907
Pryor. WA. (1976). Free radical reactions in biology: Initiation of lipid autoxidation by ozone and nitrogen
dioxide. Environ Health Perspect 16: 180-181.
Pryor. WA. (1992). How far does ozone penetrate into the pulmonary air/tissue boundary before it reacts? Free
Radic Biol Med 12: 83-88. http://dx.doi.org/10.1016/0891-5849(92)90060-T
Pryor. WA. (1994). Mechanisms of radical formation from reactions of ozone with target molecules in the
lung. Free Radic Biol Med 17: 451-465.
Pryor. WA; Bermudez. E; Cueto. R: Squadrito. GL. (1996). Detection of aldehydes in bronchoalveolar lavage
of rats exposed to ozone. Toxicol Sci 34: 148-156.
Pryor. WA; Das. B; Church. DF. (1991). The ozonation of unsaturated fatty acids: Aldehydes and hydrogen
peroxide as products and possible mediators of ozone toxicity. Chem Res Toxicol 4: 341-348.
Pryor. WA; Giamalva. DH; Church. DF. (1984). Kinetics of ozonation. 2. Amino acids and model compounds
in water and comparisons to rates in nonpolar solvents. J Am Chem Soc 106: 7094-7100.
Pulfer. MK; Murphy. RC. (2004). Formation of biologically active oxysterols during ozonolysis of cholesterol
present in lung surfactant. J Biol Chem 279: 26331-26338. http://dx.doi.org/10.1074/ibc.M403581200
Pulfer. MK; Taube. C: Gelfand. E; Murphy. RC. (2005). Ozone exposure in vivo and formation of biologically
active oxysterols in the lung. J Pharmacol Exp Ther 312: 256-264.
httrj://dx.doi.org/10.1124/ipet.l04.073437
Que. LG: Stiles. JV: Sundv. JS: Foster. WM. (2011). Pulmonary function, bronchial reactivity, and epithelial
permeability are response phenotypes to ozone and develop differentially in healthy humans. J Appl
Physiol 111: 679-687. http://dx.doi.org/10.1152/iapplphysiol.00337.2011
Rashba-Step. J: Tatovan. A; Duncan. R; Ann. D; Pushpa-Rehka. TR; Sevanian. A. (1997). Phospholipid
peroxidation induces cytosolic phospholipase A2 activity: Membrane effects versus enzyme
phosphorylation. Arch Biochem Biophys 343: 44-54. http://dx.doi.org/10.1006/abbi.1997.0134
Reeser. WH; Lee. GM; Taylor. A; Wang. L; Arnold. SF; Ultman. JS: Ben-Jebria. A. (2005). Uptake of ozone in
human lungs and its relationship to local physiological response. Inhal Toxicol 17: 699-707.
http://dx.doi.org/10.1080/08958370500224433
5-96
-------
Rigas. ML; Ben-Jebria. A; Ultman. JS. (1997). Longitudinal distribution of ozone absorption in the lung:
Effects of nitrogen dioxide, sulfur dioxide, and ozone exposures. Arch Environ Occup Health 52: 173-178.
Rigas. ML; Catlin. SN; Ben-Jebria. A; Ultman. JS. (2000). Ozone uptake in the intact human respiratory tract:
Relationship between inhaled dose and actual dose. J Appl Physiol 88: 2015-2022.
Ritz. B; Yu. F; Chapa. G: Fruin. S. (2000). Effect of air pollution on preterm birth among children born in
Southern California between 1989 and 1993. Epidemiology 11: 502-511.
Ritz. B; Yu. F; Fruin. S: Chapa. G: Shaw. GM; Harris. JA. (2002). Ambient air pollution and risk of birth
defects in Southern California. Am J Epidemiol 155: 17-25.
Romieu. I; Meneses. F; Ramirez. M; Ruiz. S: Padilla. RP; Sienra. JJ: Gerber. M; Grievink. L; Dekker. R;
Walda. I; Brunekreef. B. (1998b). Antioxidant supplementation and respiratory functions among workers
exposed to high levels of ozone. Am J Respir Crit Care Med 158: 226-232.
Romieu. I; Sienra-Monge. JJ: Ramirez-Aguilar. M; Moreno-Macias. H; Reyes-Ruiz. NI; Estela del Rio-
Navarro. B; Hernandez-Avila. M; London. SJ. (2004b). Genetic polymorphism of GSTM1 and antioxidant
supplementation influence lung function in relation to ozone exposure in asthmatic children in Mexico City.
Thorax 59: 8-10.
Romieu. I; Sienra-Monge. JJ: Ramirez-Aguilar. M; Tellez-Rojo. MM; Moreno-Macias. H; Reyes-Ruiz. NI; Del
Rio-Navarro. BE; Ruiz-Navarro. MX; Hatch. G: Slade. R; Hernandez-Avila. M. (2002). Antioxidant
supplementation and lung functions among children with asthma exposed to high levels of air pollutants.
Am J Respir Crit Care Med 166: 703-709.
Roux. E; Hyvelin. JM; Savineau. JP; Marthan. R. (1999). Human isolated airway contraction: Interaction
between air pollutants and passive sensitization. Am J Respir Crit Care Med 160: 439-445.
Rutkowski. JM; Santiag. LY; Ben-Jebria. A; Ultman. JS. (2011). Comparison of ozone-specific (OZAC) and
oxygen radical (ORAC) antioxidant capacity assays for use with nasal lavage fluid. Toxicol In Vitro 25:
1406-1413. http://dx.doi.0rg/10.1016/i.tiv.2011.04.008
Sagai. M; Arakawa. K; Ichinose. T; Shimojo. N. (1987). Biochemical effects on combined gases of nitrogen
dioxide and ozone: I. Species differences of lipid peroxides and phospholipids in lungs. Toxicology 46:
251-265. http://dx.doi.org/10.1016/0300-483X(87)90207-l
Samet JM; Hatch. GE; Horstman. D; Steck-Scott S: Arab. L; Bromberg. PA; Levine. M; McDonnell. WF;
Devlin. RB. (2001). Effect of antioxidant supplementation on ozone-induced lung injury in human subjects.
Am J Respir Crit Care Med 164: 819-825.
Santiago-Lopez. D; Bautista-Martinez. JA; Reyes-Hernandez. CI; Aguilar-Martinez. M; Rivas-Arancibia. S.
(2010). Oxidative stress, progressive damage in the substantia nigra and plasma dopamine oxidation, in rats
chronically exposed to ozone. Toxicol Lett 197: 193-200. http://dx.doi.Org/10.1016/i.toxlet.2010.05.020
Santiago. LY; Hann. MC: Ben-Jebria. A; Ultman. JS. (2001). Ozone absorption in the human nose during
unidirectional airflow. J Appl Physiol 91: 725-732.
Sarangapani. R; Gentry. PR; Covington. TR; Teeguarden. JG: Clewell HJ. III. (2003). Evaluation of the
potential impact of age- and gender-specific lung morphology and ventilation rate on the dosimetry of
vapors. Inhal Toxicol 15: 987-1016. http://dx.doi.org/10.1080/08958370390226350
Sathishkumar. K; Gao. X; Raghavamenon. AC: Parinandi. N; Pryor. WA; Uppu. RM. (2009). Cholesterol
secoaldehyde induces apoptosis in H9c2 cardiomyoblasts through reactive oxygen species involving
mitochondria! and death receptor pathways. Free Radic Biol Med 47: 548-558.
http://dx.doi.0rg/10.1016/i.freeradbiomed.2009.05.020
Sathishkumar. K; Murthy. SN; Uppu. RM. (2007a). Cytotoxic effects of oxysterols produced during ozonolysis
of cholesterol in murine GT1-7 hypothalamic neurons. Free Radic Res 41: 82-88.
http://dx.doi.org/10.1080/10715760600950566
Sathishkumar. K; Xi. X; Martin. R; Uppu. RM. (2007b). Cholesterol secoaldehyde, an ozonation product of
cholesterol, induces amyloid aggregation and apoptosis in murine GT1-7 hypothalamic neurons. J
Alzheimers Dis 11: 261-274.
5-97
-------
Sato. S: Shimura. S: Hirose. T; Maeda. S: Kawakami. M; Takishima. T; Kimura. S. (1980). Effects of long-
term ozone exposure and dietary vitamin E in rats. Tohoku J Exp Med 130: 117-128.
Sawyer. K; Brown. J; HazuchaM; Bennett. WD. (2007). The effect of exercise on nasal uptake of ozone in
healthy human adults. J Appl Physiol 102: 1380-1386. http://dx.doi.org/10.1152/iapplphysiol.00269.2006
Scannell. C: Chen. L; Aris. RM; Tager. I; Christian. D; Ferrando. R; Welch. B; Kelly. T; Balmes. JR. (1996).
Greater ozone-induced inflammatory responses in subjects with asthma. Am J Respir Crit Care Med 154:
24-29.
Schelegle. ES: Adams. WC: Siefkin. AD. (1987). Indomethacin pretreatment reduces ozone-induced
pulmonary function decrements in human subjects. Am Rev Respir Dis 136: 1350-1354.
Schelegle. ES: Carl. ML; Coleridge. HM; Coleridge. JCG: Green. JF. (1993). Contribution of vagal afferents to
respiratory reflexes evoked by acute inhalation of ozone in dogs. J Appl Physiol 74: 2338-2344.
Schelegle. ES: Miller. LA; Gershwin. LJ; Fanucchi. MV; Van Winkle. LS: Gerriets. JE; Walbv. WF; Mitchell.
V; Tarkington. BK; Wong. VJ; Baker. GL; Pantle. LM; Joad. JP; Pinkerton. KE; Wu. R; Evans. MJ; Hyde.
DM; Plopper. CG. (2003). Repeated episodes of ozone inhalation amplifies the effects of allergen
sensitization and inhalation on airway immune and structural development in Rhesus monkeys. Toxicol
Appl Pharmacol 191: 74-85.
Schelegle. ES: Siefkin. AD; McDonald. RJ. (1991). Time course of ozone-induced neutrophilia in normal
humans. Am J Respir Crit Care Med 143: 1353-1358.
Schelegle. ES: Walbv. WF; Adams. WC. (2007). Time course of ozone-induced changes in breathing pattern in
healthy exercising humans. J Appl Physiol 102: 688-697.
http://dx.doi.org/10.1152/iapplphvsiol.00141.2006
Schlesinger. RB; Ben-Jebria. A; Dahl. AR; Snipes. MB; Ultman. J. (1997). Disposition of inhaled toxicants.
In EJ Massaro (Ed.), Handbook of human toxicology (pp. 493-550). Boca Raton, FL: CRC Press.
Schroter. RC: Sudlow. MF. (1969). Flow patterns in models of the human bronchial airways. Respir Physiol
Neurobiol 7: 341-355.
Seltzer. J: Bigbv. BG: Stulbarg. M; Holtzman. MJ: Nadel. JA; Ueki. IF; Leikauf. GD; Goetzl. EJ: Boushev.
HA. (1986). 03-induced change in bronchial reactivity to methacholine and airway inflammation in
humans. J Appl Physiol 60: 1321-1326.
Servais. S: Boussouar. A; Molnar. A; Douki. T; Pequignot. JM; Favier. R. (2005). Age-related sensitivity to
lung oxidative stress during ozone exposure. Free Radic Res 39: 305-316.
http://dx.doi.org/10.1080/10715760400011098
Sharkhuu. T; Doerfler. PL; Copeland. C: Luebke. RW: Gilmour. MI. (2011). Effect of maternal exposure to
ozone on reproductive outcome and immune, inflammatory, and allergic responses in the offspring. J
Immunotoxicol 8: 183-194. http://dx.doi.org/10.3109/1547691X.2011.568978
Shore. SA. (2007). Obesity and asthma: Lessons from animal models [Review]. J Appl Physiol 102: 516-528.
http://dx.doi.org/10.1152/iapplphysiol.00847.2006
Shore. SA; Lang. JE; Kasahara. PI; Lu. FL; Verbout. NG: Si. H; Williams. ES: Terry. RD; Lee. A; Johnston.
RA. (2009). Pulmonary responses to subacute ozone exposure in obese vs. lean mice. J Appl Physiol 107:
1445-1452. http://dx.doi.org/10.1152/iapplphvsiol.00456.2009
Shore. SA; Schwartzman. IN; Le Blanc. B; Krishna Murthy. GG: Doerschuk. CM. (2001). Tumor necrosis
factor receptor 2 contributes to ozone-induced airway hyperresponsiveness in mice. Am J Respir Crit Care
Med 164: 602-607.
Sienra-Monge. JJ: Ramirez-Aguilar. M; Moreno-Macias. H; Reyes-Ruiz. NI; Del Rio-Navarro. BE; Ruiz-
Navarro. MX; Hatch. G: Crissman. K; Slade. R: Devlin. RB; Romieu. I. (2004). Antioxidant
supplementation and nasal inflammatory responses among young asthmatics exposed to high levels of
ozone. Clin Exp Immunol 138: 317-322. http://dx.doi.Org/10.llll/j.1365-2249.2004.02606.x
5-98
-------
Slade. R; Crissman. K; Norwood. J: Hatch. G. (1993). Comparison of antioxidant substances in
bronchoalveolar lavage cells and fluid from humans, guinea pigs, and rats. Exp Lung Res 19: 469-484.
Slade. R; Highfill. JW: Hatch. GE. (1989). Effects of depletion of ascorbic acid or nonprotein sulfhydryls on
the acute inhalation toxicity of nitrogen dioxide, ozone, and phosgene. Inhal Toxicol 1: 261-271.
Slade. R: Watkinson. WP; Hatch. GE. (1997). Mouse strain differences in ozone dosimetry and body
temperature changes. Am J Physiol 272: L73-L77.
Sokol. RZ; Kraft. P; Fowler. IM; Mamet. R: Kim. E; Berhane. KT. (2006). Exposure to environmental ozone
alters semen quality. Environ Health Perspect 114: 360-365. http://dx.doi.org/10.1289/ehp.8232
Stenfors. N; Pourazar. J: Blomberg. A; Krishna. MT; Mudwav. I; Helledav. R: Kelly. FJ; Frew. AJ; Sandstrom.
X, (2002). Effect of ozone on bronchial mucosal inflammation in asthmatic and healthy subjects. Respir
Med 96: 352-358.
Sterner-Kock. A; Kock. M; Braun. R; Hyde. DM. (2000). Ozone-induced epithelial injury in the ferret is
similar to nonhuman primates. Am J Respir Crit Care Med 162: 1152-1156.
Sun. J; Koto. H; Chung. KF. (1997). Interaction of ozone and allergen challenges on bronchial responsiveness
and inflammation in sensitised guinea pigs. Int Arch Allergy Immunol 112: 191-195.
Taylor-Clark. TE; McAlexander. MA; Nassenstein. C: Sheardown. SA; Wilson. S: Thornton. J: Carr. MJ;
Undem. BJ. (2008). Relative contributions of TRPA1 and TRPV1 channels in the activation of vagal
bronchopulmonary C-fibres by the endogenous autacoid 4-oxononenal. J Physiol 586: 3447-3459.
http://dx.doi.org/10.1113/jphvsiol.2008.153585
Taylor-Clark. TE; Undem. BJ. (2010). Ozone activates airway nerves via the selective stimulation of TRPAl
ion channels. J Physiol 588: 423-433. http://dx.doi.org/10.1113/iphvsiol.2009.183301
Taylor. AB; Borhan. A; Ultman. JS. (2007). Three-dimensional simulations of reactive gas uptake in single
airway bifurcations. Ann Biomed Eng 35: 235-249. http://dx.doi.org/10.1007/sl0439-006-9195-4
Taylor. AB; Lee. GM; Nellore. K; Ben-Jebria. A; Ultman. JS. (2006). Changes in the carbon dioxide
expirogram in response to ozone exposure. Toxicol Appl Pharmacol 213: 1-9.
http://dx.doi.0rg/10.1016/i.taap.2005.09.009
Tepper. JS: Costa. PL; Fitzgerald. S: Doerfler. PL; Bromberg. PA. (1993). Role of tachykinins in ozone-
induced acute lung injury in guinea pigs. J Appl Physiol 75: 1404-1411.
Tepper. JS: Costa. PL; Lehmann. JR; Weber. MF; Hatch. GE. (1989). Unattenuated structural and biochemical
alterations in the rat lung during functional adaptation to ozone. Am J Respir Crit Care Med 140: 493-501.
Thomson. E; Kumarathasan. P; Goegan. P; Aubin. RA; Vincent. R. (2005). Differential regulation of the lung
endothelin system by urban paniculate matter and ozone. Toxicol Sci 88: 103-113.
http://dx.doi.org/10.1093/toxsci/kFi272
Thomson. E; Kumarathasan. P; Vincent. R. (2006). Pulmonary expression of preproET-1 and preproET-3
mRNAs is altered reciprocally in rats after inhalation of air pollutants. Exp Biol Med 231: 979-984.
Tran. MU; Weir. AJ: Fanucchi. MV; Rodriguez. AE; Pantle. LM; Smiley-Jewell. SM; Van Winkle. LS: Evans.
MJ; Miller. LA; Schelegle. ES: Gershwin. LJ; Hyde. DM; Plopper. CG. (2004). Smooth muscle
hypertrophy in distal airways of sensitized infant rhesus monkeys exposed to house dust mite allergen. Clin
Exp Allergy 34: 1627-1633. http://dx.doi.Org/10.llll/i.1365-2222.2004.02057.x
Trenga. CA; Koenig. JO; Williams. PV. (2001). Dietary antioxidants and ozone-induced bronchial
hyperresponsiveness in adults with asthma. Arch Environ Occup Health 56: 242-249.
Trevisani. M; Siemens. J: Materazzi. S: Bautista. DM; Nassini. R; Campi. B; Imamachi. N; Andre. E;
Patacchini. R; Cottrell. GS: Gatti. R: Basbaum. AI; Bunnett. NW: Julius. D; Geppetti. P. (2007). 4-
Hydroxynonenal, an endogenous aldehyde, causes pain and neurogenic inflammation through activation of
the irritant receptor TRPAl. PNAS 104: 13519-13524. http://dx.doi.org/10.1073/pnas.0705923104
5-99
-------
Tsujino. I; Kawakami. Y; Kaneko. A. (2005). Comparative simulation of gas transport in airway models of rat,
dog, and human. Inhal Toxicol 17: 475-485. http://dx.doi.org/10.1080/08958370590964476
U.S. EPA (U.S. Environmental Protection Agency). (1986). Air quality criteria for ozone and other
photochemical oxidants [EPA Report]. (EPA-600/8-84-020aF - EPA-600/8-84-020eF). Research Triangle
Park, NC. http://www.ntis.gov/search/product.aspx?ABBR=PB87142949
U.S. EPA (U.S. Environmental Protection Agency). (1996a). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/P-93/004AF). Research Triangle Park,NC.
U.S. EPA (U.S. Environmental Protection Agency). (2005). Guidelines for carcinogen risk assessment [EPA
Report]. (EPA/630/P-03/001F). Washington, DC. http://www.epa.gov/cancerguidelines/
U.S. EPA (U.S. Environmental Protection Agency). (2006b). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/R-05/004AF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
U.S. EPA (U.S. Environmental Protection Agency). (2009f). The U.S. Environmental Protection Agencys
strategic plan for evaluating the toxicity of chemicals [EPA Report]. (EPA/1 OO/K-09/001). Washington, DC.
http://www.epa.gov/osa/spc/toxicitvtesting/docs/toxtest strategy 032309.pdf
Ultman. JS; Anjilvel. S. (1990). Monte Carlo simulation of ozone uptake in an asymmetric lung model. In DJ
Schneck; CL Lucas (Eds.), Biofluid mechanics 3: Proceedings of the third Mid-Atlantic Conference on
Biofluid Mechanics; October; Blacksburg, VA (pp. 45-52). New York, NY: New York University Press.
Ultman. JS; Ben-Jebria. A; Arnold. SF. (2004). Uptake distribution of ozone in human lungs: Intersubject
variability in physiologic response (pp. 1-23; discussion 25-30). (ISSN 1041-5505
Ultman. JS; Ben-Jebria. A; Hu. SC. (1994). Noninvasive determination of respiratory ozone absorption:
The bolus-response method. (HEI Research Report 69). Cambridge, MA: Health Effects Institute.
Uppu. RM; Cueto. R; Squadrito. GL; Pryor. WA. (1995). What does ozone react with at the air/lung interface?
Model studies using human red blood cell membranes. Arch Biochem Biophys 319: 257-266.
Urch. B; Speck. M; Corey. P; Wasserstein. D; Manno. M; Lukic. KZ; Brook. JR; Liu. L; Coull. B; Schwartz. J;
Gold. PR; Silverman. F. (2010). Concentrated ambient fine particles and not ozone induce a systemic
interleukin-6 response in humans. Inhal Toxicol 22: 210-218.
http://dx.doi.org/10.3109/08958370903173666
Vagaggini. B; Taccola. M; Clanchetti. S; Carnevali. S; Bartoli. ML; Bacci. E; Dente. FL; Di Franco. A;
Giannini. D; Paggiaro. PL. (2002). Ozone exposure increases eosinophilic airway response induced by
previous allergen challenge. Am J Respir Crit Care Med 166: 1073-1077.
Van Bree. L; Dormans. JAM. A; Koren. HS; Devlin. RB; Rombout. PJA. (2002). Attenuation and recovery of
pulmonary injury in rats following short-term, repeated daily exposure to ozone. Inhal Toxicol 14: 883-900.
http://dx.doi.org/10.1080/08958370290084674
Vancza. EM; Galdanes. K; Gunnison. A; Hatch. G; Gordon. T. (2009). Age, strain, and gender as factors for
increased sensitivity of the mouse lung to inhaled ozone. Toxicol Sci 107: 535-543.
http://dx.doi.org/10.1093/toxsci/kfn253
Vasu. VT; Oommen. S; Lim. Y; Valacchi. G; Hobson. B; Eiserich. JP; Leonard. SW; Traber. MG; Cross. CE;
Gohil. K. (2010). Modulation of ozone-sensitive genes in alpha-tocopherol transfer protein null mice. Inhal
Toxicol 22: 1-16. http://dx.doi.org/10.3109/08958370902838145
Vaughan. TR. Jr; Jennelle. LF; Lewis. TR. (1969). Long-term exposure to low levels of air pollutants: Effects
on pulmonary function in the beagle. Arch Environ Health 19: 45-50.
Verhein. KG; Hazari. MS; Moulton. BC; Jacobv. IW; Jacobv. DB; Fryer. AD. (2011). Three days after a single
exposure to ozone the mechanism of airway hyperreactivity is dependent upon substance P and nerve
growth factor. Am J Physiol Lung Cell Mol Physiol 300: L176-L184.
http://dx.doi.org/10.1152/aiplung.00060.2010
5-100
-------
Veselv. KR; Schelegle. ES: Stovall. MY; Harkema. JR: Green. JF; Hyde. DM. (1999). Breathing pattern
response and epithelial labeling in ozone-induced airway injury in neutrophil-depleted rats. Am J Respir
Cell Mol Biol 20: 699-709.
Vincent. R; Vu. D; Hatch. G: Poon. R; Dreher. K; Guenette. J: Bjarnason. S: Potvin. M; Norwood. J:
Mcmullen. E. (1996b). Sensitivity of lungs of aging Fischer 344 rats to ozone: Assessment by
bronchoalveolar lavage. Am J Physiol 271: L555-L565.
Vivier. E; Raulet. DH; Moretta. A; Caligiuri. MA; Zitvogel. L; Lanier. LL; Yokoyama. WM; Ugolini. S.
(2011). Innate or adaptive immunity? The example of natural killer cells [Review]. Science 331: 44-49.
http://dx.doi.org/10.1126/science.1198687
Vovnow. JA; Fischer. BM; Zheng. S: Potts. EN; Grover. AR; Jaiswal. AK; Ohio. AJ; Foster. WM. (2009).
NAD(P)H quinone oxidoreductase 1 is essential for ozone-induced oxidative stress in mice and humans.
Am J Respir Cell Mol Biol 41: 107-113. http://dx.doi.org/10.1165/rcmb.2008-0381OC
Wagner. JG: Harkema. JR: Jiang. Q; Illek. B; Ames. BN; Peden. DB. (2009). Gamma-tocopherol attenuates
ozone-induced exacerbation of allergic rhinosinusitis in rats. Toxicol Pathol 37: 481-491.
http://dx.doi.org/10.1177/0192623309335630
Wagner. JG: Hotchkiss. JA; Harkema. JR. (2002). Enhancement of nasal inflammatory and epithelial responses
after ozone and allergen coexposure in brown Norway rats. Toxicol Sci 67: 284-294.
Wagner. JG: Jiang. Q; Harkema. JR: Illek. B; Patel. DP; Ames. BN; Peden. DB. (2007). Ozone enhancement
of lower airway allergic inflammation is prevented by gamma-tocopherol. Free Radic Biol Med 43: 1176-
1188. http://dx.doi.0rg/10.1016/i.freeradbiomed.2007.07.013
Watkinson. WP; Campen. MJ; Nolan. JP; Costa. PL. (2001). Cardiovascular and systemic responses to inhaled
pollutants in rodents: Effects of ozone and paniculate matter. Environ Health Perspect 109: 539-546.
Watkinson. WP: Campen. MJ: Wichers. LB; Nolan. JP; Costa. PL. (2003). Cardiac and thermoregulatory
responses to inhaled pollutants in healthy and compromised rodents: Modulation via interaction with
environmental factors. Environ Res 92: 35-47.
Weibel. ER. (1980). Pesign and structure of the human lung. In AP Fishman (Ed.), Assessment of pulmonary
function (pp. 18-65). New York, NY: McGraw-Hill.
Weinmann. GG: Liu. MC: Proud. P; Weidenbach-Gerbase. M; Hubbard. W: Frank. R. (1995b). Ozone
exposure in humans: Inflammatory, small and peripheral airway responses. Am J Respir Crit Care Med
152: 1175-1182.
Weinmann. GG: Weidenbach-Gerbase. M; Foster. WM; Zacur. H; Frank. R. (1995c). Evidence for ozone-
induced small-airway dysfunction: Lack of menstrual-cycle and gender effects. Am J Respir Crit Care Med
152: 988-996.
Welch. RW: Wang. Y; Grossman. A. Jr; Park. JB; Kirk. KL; Levine. M. (1995). Accumulation of vitamin C
(ascorbate) and its oxidized metabolite dehydroascorbic acid occurs by separate mechanisms. J Biol Chem
270: 12584-12592. http://dx.doi.org/10.1074/ibc.270.21.12584
Wenten. M; Gauderman. WJ: Berhane. K; Lin. PC: Peters. J: Gilliland. FP. (2009). Functional variants in the
catalase and myeloperoxidase genes, ambient air pollution, and respiratory-related school absences:
An example of epistasis in gene-environment interactions. Am J Epidemiol 170: 1494-1501.
http://dx.doi.org/10.1093/aie/kwp310
Wiester. MJ: Stevens. MA; Menache. MG: McKee. JL. Jr; Gerritv. TR. (1996a). Ozone uptake in healthy adult
males during quiet breathing. Toxicol Sci 29: 102-109.
Wiester. MJ: Tepper. JS: King. ME; Menache. MG: Costa. PL. (1988). Comparative study of ozone (O3)
uptake in three strains of rats and in the guinea pig. Toxicol Appl Pharmacol 96: 140-146.
Wiester. MJ: Tepper. JS: Winsett. PW: Crissman. KM; Richards. JH; Costa. PL. (1996b). Adaptation to ozone
in rats and its association with ascorbic acid in the lung. Toxicol Sci 31: 56-64.
5-101
-------
Wiester. MJ; Williams. TB; King. ME; Menache. MG: Miller. FJ. (1987). Ozone uptake in awake Sprague-
Dawley rats. Toxicol Appl Pharmacol 89: 429-437. http://dx.doi.org/10.1016/0041-008X(87)90162-l
Williams. AS: Eynott. PR; Leung. SY; Natfa. P; Jupp. R; De Sanctis. GT; Resnick. R; Adcock. IM; Chung. KF.
(2009a). Role of cathepsin S in ozone-induced airway hyperresponsiveness and inflammation. Pulm
Pharmacol Ther 22: 27-32. http://dx.doi.Org/10.1016/i.pupt.2008.ll.002
Williams. AS: Issa. R: Durham. A; Leung. SY; Kapoun. A; Medicherla. S: Higgins. LS: Adcock. IM; Chung.
KF. (2008a). Role of p38 mitogen-activated protein kinase in ozone-induced airway hyperresponsiveness
and inflammation. Eur J Pharmacol 600: 117-122. http://dx.doi.Org/10.1016/i.eiphar.2008.09.031
Williams. AS: Issa. R: Leung. SY; Puneeta. N; Gregory. D; Ferguson. D; Brydon. L; Bennett. I; Adcock. M;
Chung. KF. (2007a). Attenuation of ozone-induced airway inflammation and hyper-responsiveness by c-Jun
NH2 terminal kinase inhibitor SP600125. J Pharmacol Exp Ther 322: 351-359.
http://dx.doi.org/10.1124/ipet.107.121624
Williams. AS: Leung. SY; Natfa. P; Khorasani. NM; Bhavsar. P; Issa. R: Mitchell. JA; Adcock. IM; Chung. KF.
(2007b). Role of TLR2, TLR4, and MyD88 in murine ozone-induced airway hyperresponsiveness and
neutrophilia. JAppl Physiol 103: 1189-1195. http://dx.doi.org/10.1152/iapplphvsiol.00172.2007
Williams. AS: Nath. P; Leung. SY; Khorasani. N; Mckenzie. ANJ; Adcock. IM; Chung. KF. (2008b).
Modulation of ozone-induced airway hyperresponsiveness and inflammation by interleukin-13. Eur Respir
J 32: 571-578. http://dx.doi.org/10.1183/09031936.00121607
Wu. W: Doreswamy. V; Diaz-Sanchez. D; Samet. JM; Kesic. M; Dailev. L; Zhang. W: Jaspers. I; Peden. DB.
(2011). GSTM1 modulation of IL-8 expression in human bronchial epithelial cells exposed to ozone. Free
Radic Biol Med 51: 522-529. http://dx.doi.Org/10.1016/i.freeradbiomed.2011.05.006
Wu. ZX; Barker. JS: Batchelor. TP; Dev. RD. (2008c). Interleukin (IL)-l regulates ozone-enhanced trachea!
smooth muscle responsiveness by increasing substance P (SP) production in intrinsic airway neurons of
ferret. Respir Physiol Neurobiol 164: 300-311. http://dx.doi.Org/10.1016/i.resp.2008.07.019
Wu. ZX; Satterfield. BE; Dev. RD. (2003). Substance P released from intrinsic airway neurons contributes to
ozone-enhanced airway hyperresponsiveness in ferret trachea. JAppl Physiol 95: 742-750.
Yang. IA; Holz. O; Jorres. RA; Magnussen. H; Barton. SJ: Rodriguez. S: Cakebread. JA; Hollowav. JW:
Holgate. ST. (2005a). Association of tumor necrosis factor alpha polymorphisms and ozone-induced change
in lung function. Am J Respir Crit Care Med 171: 171-176.
Yokoyama. E; Frank. R. (1972). Respiratory uptake of ozone in dogs. Arch Environ Occup Health 25: 132-
138.
Yoon. HK; Cho. HY; Kleeberger. SR. (2007). Protective role of matrix metalloproteinase-9 in ozone-induced
airway inflammation. Environ Health Perspect 115: 1557-1563. http://dx.doi.org/10.1289/ehp.10289
Yost. BL; Gleich. GJ: Jacobv. DB; Fryer. AD. (2005). The changing role of eosinophils in long-term
hyperreactivity following a single ozone exposure. Am J Physiol Lung Cell Mol Physiol 289: L627-L635.
http://dx.doi.org/10.1152/aiplung.00377.2004
5-102
-------
6 INTEGRATED HEALTH EFFECTS OF SHORT-TERM
OZONE EXPOSURE
6.1 Introduction
This chapter reviews, summarizes, and integrates the evidence for various health
outcomes associated with short-term (i.e., hours, days, or weeks) exposures to O3.
Numerous controlled human exposure, epidemiologic, and toxicological studies have
permitted evaluation of the relationships between short-term O3 exposure and a range
of endpoints related to respiratory effects (Section 6.2). cardiovascular effects
(Section 6.3). and mortality (Section 6.2. Section 6.3. and Section 6.6). A smaller
number of studies were available to assess the effects of O3 exposure on other
physiological systems such as the central nervous system (Section 6.4). liver and
metabolism (Section 6.5.1). and cutaneous and ocular tissues (Section 6.5.2). This
chapter evaluates the majority of recent [i.e., published since the completion of the
2006 O3 AQCD (U.S. EPA. 2006b)] short-term exposure studies; however, those for
birth outcomes and infant mortality are evaluated in Chapter 7 (Section 7.4). because
they compare associations among overlapping short- and long-term exposure
windows that are difficult to distinguish.
Within each individual section of this chapter, a brief summary of conclusions from
the 2006 O3 AQCD is included along with an evaluation of recent evidence that is
intended to build upon the body of evidence from previous reviews. The studies
evaluated are organized by health endpoint (e.g., lung function, pulmonary
inflammation) then by scientific discipline (e.g., controlled human exposure,
epidemiology, and toxicology). Each major section (e.g., respiratory, cardiovascular,
mortality) concludes with an integrated summary of the findings and a conclusion
regarding causality based upon the framework described in the Preamble to this ISA.
The causal determinations are presented for a broad health effect category, such as
respiratory effects, with coherence and plausibility based on the total evidence
available across disciplines and across the suite of related health endpoints, including
cause-specific mortality.
6.2 Respiratory Effects
Based on evidence integrated across controlled human exposure, epidemiologic, and
toxicological studies, the 2006 O3 AQCD concluded "that acute O3 exposure is
causally associated with respiratory system effects" (U.S. EPA. 2006b). Contributing
to this conclusion were the consistency and coherence across scientific disciplines for
the effects of short-term O3 exposure on a variety of respiratory outcomes including
"pulmonary function decrements, respiratory symptoms, lung inflammation, and
increased lung permeability, airway hyperresponsiveness." Collectively, these
findings provided biological plausibility for associations in epidemiologic studies
6-1
-------
observed between short-term increases in ambient O3 concentration and increases in
respiratory symptoms and respiratory-related hospitalizations and emergency
department (ED) visits.
Controlled human exposure studies have provided strong and quantifiable exposure-
response data on the human health effects of O3. The most salient observations from
studies reviewed in the 1996 and 2006 O3 AQCDs (U.S. EPA. 2006b. 1996a)
included: (1) young healthy adults exposed to O3 concentrations > 80 ppb develop
significant reversible, transient decrements in pulmonary function and symptoms of
breathing discomfort if minute ventilation (VE) or duration of exposure is increased
sufficiently; (2) relative to young adults, children experience similar spirometric
responses but lower incidence of symptoms from O3 exposure; (3) relative to young
adults, O3-induced spirometric responses are decreased in older individuals; (4) there
is a large degree of intersubject variability in physiologic and symptomatic responses
to O3, but responses tend to be reproducible within a given individual over a period
of several months; (5) subjects exposed repeatedly to O3 for several days experience
an attenuation of spirometric and symptomatic responses on successive exposures,
which is lost after about a week without exposure; and (6) acute O3 exposure initiates
an inflammatory response that may persist for at least 18 to 24 hours postexposure.
Substantial evidence for biologically plausible O3-induced respiratory morbidity has
been derived from the coherence between toxicological and controlled human
exposure study findings for parallel endpoints. For example, O3-induced lung
function decrements and increased airway hyperresponsiveness have been observed
in both animals and humans. Airway hyperresponsiveness could be an important
consequence of exposure to ambient O3 because the airways are then predisposed to
narrowing upon inhalation of a variety of ambient stimuli. Additionally, airway
hyperresponsiveness tends to resolve more slowly and appears less subject to
attenuation with repeated O3 exposures than lung function decrements. Increased
permeability and inflammation have been observed in the airways of humans and
animals alike after O3 exposure, although these processes are not necessarily
associated with immediate changes in lung function or hyperresponsiveness.
Furthermore, the potential relationship between repetitive bouts of acute
inflammation and the development of chronic respiratory disease is unknown.
Another feature of O3-related respiratory morbidity is impaired host defense and
reduced resistance to lung infection, which has been strongly supported by
toxicological evidence and, to a limited extent, by evidence from controlled human
exposure studies. Recurrent respiratory infection in early life is associated with
increased incidence of asthma in humans.
In concordance with experimental studies, epidemiologic studies have provided clear
evidence for decrements in lung function related to short-term ambient O3 exposure.
These effects have been demonstrated in healthy children attending camps, adults
exercising or working outdoors, and children with and without asthma (U.S. EPA.
2006K 1996a). In addition to lung function decrements, short-term increases in
ambient O3 concentration have been associated with increases in respiratory
symptoms (e.g., cough, wheeze, shortness of breath), notably in large U.S. panel
6-2
-------
studies of children with asthma (Gent et al., 2003; Mortimer et al., 2000).
The evidence across disciplines for O3 effects on a range of respiratory endpoints
collectively provides support for epidemiologic studies that have demonstrated
consistent associations between short-term increases in ambient O3 concentration and
increases in respiratory hospital admissions and ED visits, specifically during the
summer or warm months. In contrast with other respiratory health endpoints,
previous epidemiologic evidence has not clearly supported a relationship between
short-term O3 exposure and respiratory mortality. Although O3 has been consistently
associated with nonaccidental and cardiopulmonary mortality, the contribution of
respiratory causes to these findings was uncertain as the few studies that have
examined mortality specifically from respiratory causes reported inconsistent
associations with ambient O3 concentrations.
As will be discussed throughout this section, consistent with the strong body of
evidence presented in the 2006 O3 AQCD, recent studies continue to support
associations between short-term O3 exposure and respiratory effects, in particular,
lung function decrements in controlled human exposure studies, airway inflammatory
responses in toxicological studies, and respiratory-related hospitalizations and ED
visits. Recent epidemiologic studies contribute new evidence for potentially at-risk
populations and associations linking ambient O3 concentrations with biological
markers of airway inflammation and oxidative stress, which is consistent with the
extensive evidence from controlled human exposure and toxicological studies.
Furthermore, extending the potential range of well-established O3-associated
respiratory effects, recent multicity studies and a multicontinent study demonstrate
associations between short-term increases in ambient O3 concentration and
respiratory-related mortality.
6.2.1 Lung Function
6.2.1.1 Controlled Human Exposure
This section focuses on studies examining O3 effects on lung function and
respiratory symptoms in volunteers exposed, for periods of up to 8 hours, to
O3 concentrations ranging from 40 to 500 ppb, while at rest or during exercise of
varying intensity. Responses to acute O3 exposures in the range of ambient
concentrations include decreased inspiratory capacity; mild bronchoconstriction;
rapid, shallow breathing patterns during exercise; and symptoms of cough and pain
on deep inspiration (PDI). Reflex inhibition of inspiration results in a decrease in
forced vital capacity (FVC) and total lung capacity (TLC) and, in combination with
mild bronchoconstriction, contributes to a decrease in the forced expiratory volume
in 1 second (FEVi).
In studies that have exposed subjects during exercise, the majority of shorter duration
(< 4-hour exposures) studies utilized an intermittent exercise protocol in which
6-3
-------
subjects rotated between 15-minute periods of exercise and rest. A limited number of
1- to 2-hour studies, mainly focusing on exercise performance, have utilized a
continuous exercise regime. A quasi continuous exercise protocol is common to
prolonged exposure studies where subjects complete 50-minute periods of exercise
followed by 10-minute rest periods.
The majority of controlled human exposure studies have been conducted within
exposure chambers, although a smaller number of studies used a facemask to expose
subjects to O3. Little effort has been made herein to differentiate between facemask
and chamber exposures since FEVi and respiratory symptom responses appear
minimally differentially affected by these exposure modalities. Similar responses
between facemask and chamber exposures have been reported for exposures to 80
and 120 ppb O3 (6.6-hour, moderate quasi continuous exercise, 40 L/min) and
300 ppb O3 (2 hours, heavy intermittent exercise, 70 L/min) (Adams, 2003a, b,
2002).
The majority of controlled human exposure studies investigating the effects O3 are of
a randomized, controlled, crossover design in which subj ects were exposed, without
knowledge of the exposure condition and in random order to clean filtered air (FA;
the control) and, depending on the study, to one or more O3 concentrations. The FA
control exposure provides an unbiased estimate of the effects of the experimental
procedures on the outcome(s) of interest. Comparison of responses following this FA
exposure to those following an O3 exposure allows for estimation of the effects of O3
itself on an outcome measurement while controlling for independent effects of the
experimental procedures. As individuals may experience small changes in various
health endpoints from exercise, diurnal variation, or other effects in addition to those
of O3 during the course of an exposure, the term "O3-induced" is used herein to
designate effects that have been corrected or adjusted for such extraneous responses
as measured during FA exposures.
Spirometry, viz., FEVi, is a common health endpoint used to assess effects of O3 on
respiratory health in controlled human exposure studies. In considering 6.6-hour
exposures to FA, group mean FEVi changes have ranged from -0.7% (McDonnell et
al.. 1991) to 2.7% (Adams. 2006a). On average, across ten 6.6-hour exposure studies,
there has been a 1.0% (n = 279) increase in FEVi (Kim et al.. 2011: Schelegle et al..
2009: Adams. 2006a. 2003a. 2002: Adams and Ollison. 1997: Folinsbee et al.. 1994:
McDonnell et al.. 1991: Horstman et al.. 1990: Folinsbee et al.. 1988). Regardless of
the reason for small changes in FEVi over the course of FA exposures, whether
biologically based or a systematic effect of the experimental procedures, the use of
FA responses as a control for the assessment of responses following O3 exposure in
randomized exposure studies serves to eliminate alternative explanations other than
those of O3 itself in causing the measured responses.
With respect to FEVi responses in young healthy adults, an O3-induced change in
FEVi is typically the difference between the decrement observed with O3 exposure
and the improvement observed with FA exposure. Noting that some healthy
individuals experience small improvements while others have small decrements in
FEVi following FA exposure, investigators have used the randomized, crossover
6-4
-------
design with each subject serving as their own control (exposure to FA) to discern
relatively small effects with certainty since alternative explanations for these effects
are controlled for by the nature of the experimental design. The utility of
intraindividual FA control exposures becomes more apparent when considering
individuals with respiratory disease. The occurrence of exercise-induced
bronchospasm is well recognized in patients with asthma and chronic obstructive
pulmonary disease (COPD) and may be experienced during both FA and O3
exposures. Absent correction for FA responses, exercise-induced changes in FEVi
could be mistaken for responses due to O3. This biological phenomenon serves as an
example to emphasize the need for a proper control exposure in assessing the effects
of O3 as well as the role of this control in eliminating the influence of other factors
on the outcomes of interest.
Pulmonary Function Effects of Ozone Exposure in Healthy Subjects
Acute Exposure of Healthy Subjects
The majority of controlled human exposure studies have investigated the effects of
exposure to O3 in young healthy nonsmoking adults (18-35 years of age). These
studies typically use fixed concentrations of O3 under carefully regulated
environmental conditions and subject activity levels. The magnitude of respiratory
effects (decrements in spirometry measurements and increases in symptomatic
responses) in these individuals is a function of O3 concentration (C), minute
ventilation (VE), and exposure duration (time). Any physical activity will increase
minute ventilation and therefore the dose of inhaled O3. Dose of inhaled O3 to the
lower airways is also increased due to a shift from nasal to oronasal breathing with a
consequential decrease in O3 scrubbing by the upper airways. Thus, the intensity of
physiological response following an acute exposure will be strongly associated with
minute ventilation.
The product of C x VE x time is commonly used as a surrogate for O3 dose to the
respiratory tract in controlled human exposure studies. A large body of data
regarding the interdependent effects of C, VE, and time on pulmonary responses was
assessed in the 1986 and 1996 O3 AQCDs (U.S. EPA. 1996a. 1986). Acute responses
were modeled as a function of total inhaled dose (C x VE x time) which was found to
be a better predictor of response to O3 than C, VE, or time of exposure, alone, or as a
combination of any two of these factors. However, intake dose (C x VE x time) did
not adequately capture the temporal dynamics of pulmonary responses in a
comparison between a constant (square-wave) and a variable (triangular) O3
exposure (average 120 ppb O3; moderate exercise, VE = 40 L/min; 8 hour duration)
conducted by Hazucha et al. (1992). Recent nonlinear statistical models clearly
describe the temporal dynamics of FEVi responses as a function of C, VE, time, and
age of the exposed subject (McDonnell et al.. 2010. 2007).
For healthy young adults exposed at rest for 2 hours, 500 ppb is the lowest O3
concentration reported to produce a statistically significant O3-induced group mean
6-5
-------
FEVi decrement of 6.4% (n = 10) (Folinsbee et al.. 1978) to 6.7% (n = 13) (Horvath
et al., 1979). Airway resistance was not clearly affected during at-rest exposure to
these O3 concentrations. For exposures of 1-2 hours to > 120 ppb O3, statistically
significant symptomatic responses and effects on FEVi are observed when VE is
sufficiently increased by exercise (McDonnell et al., 1999b). For instance, 5% of
young healthy adults exposed to 400 ppb O3 for 2 hours during rest experienced pain
on deep inspiration. Respiratory symptoms were not observed at lower exposure
concentrations (120-300 ppb) or with only 1 hour of exposure even at 400 ppb.
However, when exposed to 120 ppb O3 for 2 hours during light-to-moderate
intermittent exercise (VE of 22 - 35 L/min), 9% of individuals experienced pain on
deep inspiration, 5% experienced cough, and 4% experienced shortness of breath.
With very heavy continuous exercise (VE = 89 L/min), an O3-induced group mean
decrement of 9.7% in FEVi has been reported for healthy young adults exposed for
1 hour to 120 ppb O3 (Gong et al., 1986). Symptoms are present and decrements in
forced expiratory volumes and flows occur at 160-240 ppb O3 following 1 hour of
continuous heavy exercise (VE « 55 to 90 L/min (Gong et al., 1986; Avol et al., 1984;
Folinsbee et al., 1984; Adams and Schelegle, 1983) and following 2 hours of
intermittent heavy exercise (VE « 65-68 L/min) (Linn et al.. 1986; Kulle et al.. 1985;
McDonnell et al.. 1983). With heavy intermittent exercise (15-min intervals of rest
and exercise [VE = 68 L/min]), symptoms of breathing discomfort and a group mean
O3-induced decrement of 3.4% in FEVi occurred in young healthy adults exposed
for 2 hours to 120 ppb O3 (McDonnell et al.. 1983).1 Table 6-1 provides examples of
typical exercise protocols utilized in controlled human exposures to O3. The VE rates
in this table are per body surface area (BSA) which is, on average, about 1.7 m2 and
2.0 m2 for young healthy adult females and males, respectively, who participated in
controlled O3 exposure studies.
1 In total, subjects were exposed to O3 for 2.5 hours. Intermittent exercise periods, however, were only conducted for the first
2 hours of exposure and FEVi was determined 5 minutes after the exercise was completed.
6-6
-------
Table 6-1 Activity levels used in controlled exposures of healthy young
adults to O3.
Activity3'13
Rest
Light quasi-continuous
exercise
Moderate quasi-
continuous exercise
Heavy intermittent
exercise
Very heavy continuous
exercise
Study
Duration
(hours)
2
6.6-7.6
6.6
1-2
1
VE
(L/min per
m2 BSA)
4
15
17-23
27-33
45
Heart
Rate
(bpm)
70
110
115-130
160
160
Treadmill
Speed (mph)
n.a.
3.5-4.4
3.3-3.5
3.5-5
n.a.
Treadmill
Grade (%)
n.a.
0
4-5
10-12
n.a.
Cycle
(watts)
n.a.
42
72
100
260
"Based on group mean exercise specific data provided in the individual studies. On average, subjects were 23 years of age. For
exercise protocols, the minute ventilation and heart rate are for the exercise periods. Quasi-continuous exercise consists of 50
minutes of exercise periods followed by 10 minutes of rest. Intermittent exercise consists of alternating periods of 15 minutes of
exercise and 15 minutes of rest.
""References: Horvath et al. (1979) for rest; Adams (2000) and Horstman et al. (1995) for light quasi-continuous exercise; Adams
(2006a. 2002. 2000). Folinsbee etal. (1988). Horstman et al. (1990). and McDonnell et al. (1991) for moderate quasi-continuous
exercise; Kehrlet al. (1987). Kreit etal. (1989). and McDonnell etal. (1983) for heavy intermittent exercise, and Gong etal. (1986)
for very heavy continuous exercise.
6-7
-------
20 i
•0 Ł. 15 -
o —
o c
3 a>
C fl>
"7 o 10 -
w =
u_ i>
* Adams (2006J
A Adams (2003)
* Adams (2002)
° Folinsbee et al (1988)
D Horslman etal. (1990)
O McDonnell etal. (1991)
McDonnell el al (200?)
0.02 0-04
0.06 0.08 0.1
Ozone (ppm)
0.12
0.14
Source: Brown et al. (2008).
8% n
— 6% -
0) *-
o c
3 0)
•o E
•T Ł
II
4% -
^T 2% -
* Adams (2006)
A Adams (2003)
X Adams (2002)
D Horstmanetal.(1990)
O Kim etal. (2011)
C McDonnell etal. (1991)
ASchelegleetal.(2009)
* (t)
A (t)
(m)
A (t)
A (t,m)
*A (t)
** (t)
O
X (m)
*(t)
0.03 0.04 0.05
0.06
0.07
0.08 0.0
B
Ozone (ppm)
Top, panel A: All studies exposed subjects to a constant (square-wave) concentration in a chamber, except Adams (2002) where a
facemask was used. All responses at and above 0.06 ppm were statistically significant. The McDonnell et al. (2007) curve
illustrates the predicted FEVi decrement at 6.6 hours as a function of O3 concentration for a 23 year-old (the average age of
subjects that participated in the illustrated studies). Note that this curve was not "fitted" to the plotted data. Error bars (where
available) are the standard error of responses.
Bottom, panel B: All studies used constant (square-wave) exposures in a chamber unless designated as triangular (t) and/or
facemask (m) exposures. All responses at and above 0.07 ppm were statistically significant. At 0.06 ppm, Adams (2006a) and Kim
et al. (2011) responses to square-wave chamber exposures were statistically significant. During each hour of the exposures,
subjects were engaged in moderate quasi continuous exercise (40 L/min) for 50 minutes and rest for 10 minutes. Following the
third hour, subjects had an additional 35-minute rest period for lunch. The data at 0.06, 0.08 and 0.12 ppm have been offset for
illustrative purposes.
Studies appearing in the figure legends: Adams (2006a. 2003a. 2002). Folinsbee et al. (1988). Horstman et al. (1990). Kim et al.
(2011). McDonnell et al. (2007): McDonnell et al. (1991). and Schelegle et al. (2009).
Figure 6-1 Cross-study comparison of mean O3-induced FEVi decrements
following 6.6 hours of exposure to O3.
6-8
-------
For prolonged (6.6 hours) exposures relative to shorter exposures, significant
pulmonary function responses and symptoms have been observed at lower
O3 concentrations and at a moderate level of exercise (VE = 40 L/min). The 6.6-hour
experimental protocol was intended to simulate the performance of heavy physical
labor for a full workday (Folinsbee et al.. 1988). The results from studies using
6.6 hours of constant or square-wave exposures to between 40 and 120 ppb O3 are
illustrated in Figure 6-1 (A). Figure 6-l(B) focuses on the range from 40 to 80 ppb
and includes triangular exposure protocols as well as facemask exposures. Exposure
to 40 ppb O3 for 6.6 hours produces small, statistically nonsignificant changes in
FEVi that are relatively similar to responses from FA exposure (Adams, 2002).
Volunteers exposed to 60 ppb O3 experience group mean O3-induced FEVi
decrements of about 3% (Kim et al.. 2011: Brown et al.. 2008: Adams. 2006aV:
those exposed to 80 ppb have group mean decrements that range from 6 to 8%
(Adams. 2006a. 2003a: McDonnell et al.. 1991: Horstman et al.. 1990): at 100 ppb,
group mean decrements range from 8 to 14% (McDonnell et al.. 1991: Horstman et
al.. 1990): and at 120 ppb, group mean decrements of 13 to 16% are observed
(Adams. 2002: Horstman et al.. 1990: Folinsbee et al.. 1988). As illustrated in
Figure 6-1. there is a smooth intake dose-response curve without evidence of a
threshold for exposures between 40 and 120 ppb O3. This is consistent with Hazucha
and Lefohn (2007). who suggested that a randomly selected group of healthy
individuals of sufficient size would include hypo-, normo-, and hyper-responsive
individuals such that the average response would show no threshold for any
spirometric endpoint. Taken together, these data indicate that mean FEVi is clearly
decreased by 6.6-hour exposures to 60 ppb O3 and higher concentrations in subjects
performing moderate exercise. Discussed later in this subsection, the recent
McDonnell et al. (2012) and Schelegle et al. (2012) studies analyzed large datasets
and fit compartmental models, which included the concept of a response threshold.
The time course of responses during prolonged (6.6 hours) square-wave O3
exposures with moderate exercise (VE = 40 L/min) depends on O3 concentration.
At 120 ppb O3, Folinsbee et al. (1988) observed that somewhat small FEVi
decrements and symptoms of breathing discomfort become apparent in healthy
subjects following the second hour of exposure with a more rapid change in
responses between the 3rd and 5th hour of exposure and a diminishing response or
plateau in responses over the last hour of exposure. Relative to FA, the change in
FEVi at 120 ppb O3 became statistically significant after 4.6 hours. Following the
same exposure protocol, Horstman et al. (1990) observed a linear increase in FEVi
responses with time following 2 hours of exposure to 120 ppb O3 that was
statistically different from FA responses after 3 hours. At 100 ppb O3, FEVi
responses diverged from FA after 3 hours and were statistically different at 4.6 hours
(Horstman et al.. 1990). At 80 ppb O3, FEVi responses diverged from FA after 4.6
hours and were statistically different from FA at 5.6 hours (Horstman et al.. 1990).
Subsequently, Adams (2006a) observed FEVi decrements and total respiratory
1 Adams (2006a) did not find effects on FE\A| at 60 ppb to be statistically significant. In an analysis of the Adams (2006a) data, even
after removal of potential outliers, Brown et al. (2008) found the average effect on FEV, at 60 ppb to be small, but highly
statistically significant (p < 0.002) using several common statistical tests.
6-9
-------
symptoms at 80 ppb O3 to diverge from FA responses after 3 hours, but did not
become statistically different until 6.6 hours. At 60 ppb O3, FEVi responses
generally tracked responses in FA for the first 4.6 hours of exposure and diverged
after 5.6 hours (Adams, 2006a). FEVi responses, but not symptomatic responses,
become statistically different between 60 ppb O3 and FA at 6.6 hours (Kim et al,
2011: Brown et al.. 2008). At 40 ppb, FEVi and symptomatic responses track FA for
5.6 hours of exposure and may begin to diverge after 6.6 hours (Adams. 2002).
In prolonged (6.6 hours) square-wave O3 exposures between 40 and 120 ppb with
moderate exercise (VE = 40 L/min), the time required for group mean responses to
differ between O3 and FA exposures increases with decreasing O3 concentration.
As opposed to constant (i.e., square-wave) concentration patterns used in the studies
described above, many studies conducted at the levels of 40-80 ppb have used
variable (i.e., triangular) O3 concentration patterns. It has been suggested that a
triangular exposure profile can potentially lead to higher FEVi responses than
square-wave profiles despite having the same average O3 concentration over the
exposure period. Hazucha et al. (1992) were the first to investigate the effects of
variable versus constant concentration exposures on responsiveness to O3. In their
study, volunteers were randomly exposed to a triangular concentration profile
(averaging 120 ppb over the 8-hour exposure) that increased linearly from 0-240 ppb
for the first 4 hours of the 8-hour exposure, then decreased linearly from 240 to 0 ppb
over the next 4 hours of the 8-hour exposure, and to an square-wave exposure of
120 ppb O3 for 8 hours. While the total inhaled O3 doses at 4 hours and 8 hours for
the square-wave and the triangular concentration profile were almost identical, the
FEVi responses were dissimilar. For the square-wave exposure, FEVi declined ~5%
by the fifth hour and then remained at that level. With the triangular O3 profile, there
was minimal FEVi response over the first 3 hours followed by a rapid decrease in
FEVi to a decrement of 10.3% over the next 3 hours. During the seventh and eighth
hours, mean FEVi decrement improved to 6.3% as the O3 concentration decreased
from 120 to 0 ppb (mean = 60 ppb). These findings illustrate that the severity of
symptoms and the magnitude of spirometric responses are time-dependent functions
of O3 delivery rate with periods of both effect development and recovery during the
course of an exposure.
Subsequently, others have also demonstrated that variable concentration exposures
can elicit greater FEVi and symptomatic responses than do square-wave exposures to
O3 (Adams, 2006a, b, 2003 a). Adams (2006b) reproduced the findings of Hazucha et
al. (1992) at 120 ppb. However, Adams (2006a, 2003a) found that responses from an
80 ppb O3 (average) triangular exposure did not differ significantly from those
observed with the 80 ppb O3 square-wave exposure at 6.6 hours. Nevertheless, FEVi
and symptoms were significantly different from pre-exposure at 4.6 hours (when the
O3 concentration was 150 ppb) in the triangular exposure, but not until 6.6 hours in
the square-wave exposure. At the lower O3 concentration of 60 ppb, no temporal
pattern differences in FEVi responses could be discerned between square-wave and
triangular exposure profiles (Adams, 2006a). However, both total symptom scores
and pain on deep inspiration tended to be greater following the 60 ppb triangular than
the 60 ppb square-wave exposure. At 80 ppb O3, respiratory symptoms tended to
6-10
-------
increase more rapidly during the triangular than square-wave exposure protocol, but
then decreased during the last hour of exposure to be less than that observed with the
square-wave exposure at 6.6 hours. Both total symptom scores and pain on deep
inspiration were significantly increased following exposures to 80 ppb O3 relative to
all other exposure protocols, i.e., FA, 40, and 60 ppb exposures. Following the
6.6-hour exposures, respiratory symptoms at 80 ppb were roughly 2-3 times greater
than those observed at 60 ppb. At 40 ppb, triangular and square-wave patterns
produced spirometric and subjective symptom responses similar to FA exposure
(Adams. 2006a. 2002).
For O3 exposures of 60 ppb and greater, studies (Adams, 2006a, b, 2003a; Hazucha
et al., 1992) demonstrate that during triangular exposure protocols, volunteers
exposed during moderate exercise (VE = 40 L/min) may develop greater spirometric
and/or symptomatic responses during and following peak O3 concentrations as
compared to responses over the same time interval of square-wave exposures. This
observation is not unexpected since the inhaled dose rate during peaks of the
triangular protocols approached twice that of the square-wave protocols,
e.g., 150 ppb versus 80 ppb peak concentration. At time intervals toward the end of
an exposure, O3 delivery rates for the triangular protocols were less than those of
square-wave. At these later time intervals, there is some recovery of responses during
triangular exposure protocols, whereas there is a continued development of or a
plateau of responses in the square-wave exposure protocols. Thus, responses during
triangular protocols relative to square-wave protocols may be expected to diverge
and be greater following peak exposures and then converge toward the end of an
exposure. Subsequent discussion will focus on exposures between 40 and 80 ppb
where FEVi pre-to-post responses are similar (although not identical) between
triangular and square-wave protocols having equivalent average exposure
concentrations.
Schelegle et al. (2009) recently investigated the effects of 6.6-hour variable O3
exposure protocols at mean concentrations of 60, 70, 80, and 87 ppb on respiratory
symptoms and pulmonary function in young healthy adults (16 F, 15 M;
21.4 ± 0.6 years) exposed during moderate quasi continuous exercise (VE = 40
L/min). The mean FEVi (± standard error) decrements at 6.6 hours (end of exposure
relative to pre-exposure) were: -0.80 ± 0.90%, 2.72 ± 1.48%, 5.34 ± 1.42%,
7.02 ± 1.60%, and 11.42 ± 2.20% for exposure to FA, 60, 70, 80, and 87 ppb O3,
respectively. Statistically significant decrements in FEVi and increases in total
subjective symptom scores (p <0.05) were found following exposure to mean
concentrations of 70, 80, and 87 ppb O3 relative to FA. Statistically significant
effects were not found at 60 ppb. One of the expressed purposes of the Schelegle et
al. (2009) study was to determine the minimal mean O3 concentration that produces a
statistically significant decrement in FEVi and respiratory symptoms in healthy
individuals completing 6.6-hour exposure protocols. At 70 ppb, Schelegle et al.
(2009) observed a statistically significant O3-induced FEVi decrement of 6.1% at
6.6 hours and a significant increase in total subjective symptoms at 5.6 and 6.6 hours.
A re-analysis found the FEVi responses at 70 ppb to be significantly different from
FA responses beginning at 4.6 hours of exposure (Lefohn et al., 2010a). At 60 ppb,
6-11
-------
an Os-induced 3.5% FEVi decrement was not found to be statistically significant.
However, this effect is similar in magnitude to the 2.9% FEVi decrement at 60 ppb
observed by Adams (2006a). which was found to be statistically significant by
Brown et al. (2008).
More recently, Kim et al. (2011) investigated the effects of a 6.6-hour exposure to
60 ppb O3 during moderate quasi continuous exercise (VE = 40 L/min) on pulmonary
function and respiratory symptoms in young healthy adults (32 F, 27 M;
25.0 ± 0.5 years) who were roughly half glutathione S-transferase u-1 (GSTMl)-null
genetically and half GSTM1-positive. Sputum neutrophil levels were also measured
in a subset of the subjects (13 F, 11 M). The mean FEVi (± standard error)
decrements at 6.6 hours (end of exposure relative to pre-exposure) were significantly
different (p = 0.008) between the FA (0.002 ± 0.46%) and O3 (1.76 ± 0.50%)
exposures. The inflammatory response following O3 exposure was also significantly
(p O.001) increased relative to the FA exposure. Respiratory symptoms were not
affected by O3 exposure. There was also no significant effect of GSTM1 genotype on
FEVi or inflammatory responses to O3.
Consideration of the minimal O3 concentration producing statistically significant
effects on FEVi and respiratory symptoms (e.g., cough and pain on deep inspiration)
following 6.6-hour exposures warrants additional discussion. As discussed above,
numerous studies have demonstrated statistically significant O3-induced group mean
FEVi decrements of 6-8% and an increase in respiratory symptoms at 80 ppb.
Schelegle et al. (2009) have now reported a statistically significant O3-induced group
mean FEVi decrement of 6%, as well as increased respiratory symptoms, at 70 ppb.
At 60 ppb, there is information available from 4 separate studies (Kim et al.. 2011:
Schelegle et al.. 2009: Adams. 2006a. 2002).1 The group mean O3-induced FEVi
decrements observed in these studies were 3.6% (facemask, square-wave) by Adams
(2006a. 2002)2, 2.8% (triangular exposure) and 2.9% (square-wave exposure) by
Adams (2006a). 3.5% (triangular exposure) by Schelegle et al. (2009). and 1.8%
(square-wave exposure) by Kim et al. (2011). Based on data from these studies, at
60 ppb, the weighted-average group mean O3-induced FEVi decrement
(i.e., adjusted for FA responses) is 2.7% (n = 150). Although not consistently
statistically significant, these group mean changes in FEVi at 60 ppb are consistent
among studies, i.e., none observed an average improvement in lung function
following a 6.6-hour exposure to 60 ppb O3. Indeed, as was illustrated in Figure 6-1.
the group mean FEVi responses at 60 ppb fall on a smooth intake dose-response
curve for exposures between 40 and 120 ppb O3. Furthermore, in a re-analysis of the
60 ppb square-wave data from Adams (2006a). Brown et al. (2008) found the mean
effects on FEVi to be highly statistically significant (p <0.002) using several
common statistical tests even after removal of 3 potential outliers. A statistically
1 Adams (2006a): (2002) both provide data for an additional group of 30 healthy subjects that were exposed via facemask to 60 ppb
(square-wave) O3 for 6.6 hours with moderate exercise (VE = 23 L/min per m2 BSA). These subjects are described on page 133 of
Adams (2006a) and pages 747 and 761 of Adams (2002). The FEVi decrement may be somewhat increased due to a target VE of
23 L/min per m2 BSA relative to other studies with which it is listed having the target VE of 20 L/min per m2 BSA. Based on Adams
(2003a. b, 2002) the facemask exposure is not expect to affect the FENA responses relative to a chamber exposure.
2 This group average FEVi response is for a set of subjects exposed via facemask to 60 ppb O3, see page 133 of Adams (2006a).
6-12
-------
significant increase in total respiratory symptoms at 60 ppb has only been reported
by Adams (2006a) for a triangular exposure protocol at 5.6 hours and 6.6 hours
relative to baseline (not FA). Although not statistically significant, there was a
tendency for an increase in total symptoms and pain on deep inspiration following
the 60 ppb exposures (triangular and square-wave) relative to those following both
FA and 40 ppb exposures. The time-course and magnitude of FEVi responses at
40 ppb resemble those occurring during FA exposures (Adams. 2006a. 2002). In both
of these studies, there was a tendency (not statistically significant) for a small
increase in total symptoms and pain on deep inspiration following the 40 ppb
exposures relative to those following FA. Taken together, the available evidence
shows that detectable effects of O3 on group mean FEVi persist down to 60 ppb, but
not 40 ppb in young healthy adults exposed for 6.6 hours during moderate exercise.
Although group mean FEVi responses at 60 ppb are relatively small (2-3% mean
FEVi decrement), it should be emphasized that there is considerable intersubject
variability, with some responsive individuals consistently experiencing larger than
average FEVi responses.
In addition to overt effects of O3 exposure on the large airways indicated by
spirometric responses, O3 exposure also affects the function of the small airways and
parenchymal lung. Foster et al. (1997); (1993) examined the effect of O3 on
ventilation distribution. In healthy adult males (n = 6; and 26.7 ± 7 years old)
exposed to O3 (330 ppb with light intermittent exercise for 2 hours), there was a
significant reduction in ventilation to the lower lung (31% of lung volume) and
significant increases in ventilation to the upper- and middle-lung regions (Foster et
al., 1993). In a subsequent study of healthy males (n = 15; and 25.4 ± 2 years old)
exposed to O3 (350 ppb with moderate intermittent exercise for 2.2 hours), O3
exposure caused a delayed gas washout in addition to a 14% FEVi decrement (Foster
et al.. 1997). The pronounced slow phase of gas washout following O3 exposure
represented a 24% decrease in the washout rate. A day following O3 exposure, 50%
of the subjects still had (or developed) a delayed washout relative to the pre-O3
maneuver. These studies suggest a prolonged O3 effect on the small airways and
ventilation distribution in healthy young individuals.
There is a rapid recovery of O3-induced spirometric responses and symptoms; 40 to
65% recovery appears to occur within about 2 hours following exposure (Folinsbee
andHazucha, 1989). For example, following a 2-hour exposure to 400 ppb O3 with
intermittent exercise, Nightingale et al. (2000) observed a 13.5% mean decrement in
FEVi. By 3 hours postexposure, however, only a 2.7% FEVi decrement persisted.
Partial recovery also occurs following cessation of exercise despite continued
exposure to O3 (Folinsbee et al., 1977) and at low O3 concentrations during exposure
(Hazucha et al., 1992). A slower recovery phase, especially after exposure to higher
O3 concentrations, may take at least 24 hours to complete (Folinsbee and Hazucha,
2000; Folinsbee et al., 1993). Repeated daily exposure studies at higher
concentrations typically show that FEVi response to O3 is enhanced on the
second day of exposure. This enhanced response suggests a residual effect of the
previous exposure, about 22 hours earlier, even though the pre-exposure spirometry
may be the same as on the previous day. The absence of the enhanced response with
6-13
-------
repeated exposure at lower O3 concentrations may be the result of a more complete
recovery or less damage to pulmonary tissues (Folinsbee et al.. 1994).
Predicted Responses in Healthy Subjects
Studies analyzing large data sets (hundreds of subjects) provide better predictive
ability of acute changes in FEVi at low levels of O3 and VE than is possible via
comparisons between smaller studies. A few such studies described in the 2006 O3
AQCD (U.S. EPA. 2006b) analyzed FEVi responses in healthy young adults (18-
35 years of age) recruited from the area around Chapel Hill, NC and exposed for 2
hours to O3 concentrations of up to 400 ppb at rest or with intermittent exercise
(McDonnell et al.. 1997: Seal et al.. 1996: Seal et al.. 1993). McDonnell et al.
(1999b) examined changes in respiratory symptoms with O3 exposure in a subset of
the Chapel Hill data. In general, these studies showed that FEVi and respiratory
symptom responses increase with increasing O3 concentration and VE and decrease
with increasing subject age. More recent studies expand upon these analyses of FEVi
responses to also include longer duration (up to 8 hours) studies and periods of
recovery following exposure.
McDonnell et al. (2007) provided a nonlinear empirical model for predicting group
average FEVi responses as a function of O3 concentration, exposure time, VE, and
age of the exposed individual. The model predicts temporal dynamics of FEVi
change in response to any set of O3 exposure conditions that might reasonably be
experienced in the ambient environment. The model substantially differs from earlier
statistical models in that it effectively considers the concurrent processes of damage
and repair, i.e., the model allows effects on FEVi to accumulate during exposure at
the same time they are reduced due to the reversible nature of the effects. The model
was based on response data of healthy, nonsmoking, white males (n = 541), 18-
35 years old, from 15 studies conducted at the U.S. EPA Human Studies Facility in
Chapel Hill, NC.
McDonnell et al. (2010) tested the predictive ability of the model (Mcdonnell et al..
2007) against independent data (i.e., data that were not used to fit the model) of
Adams (2006a. b, 2003a. 2002. 2000). Hazucha et al. (1992). and Schelegle et al.
(2009). The model generally captured the dynamics of group average FEVi
responses within about a one percentage point of the experimental data. Consistent
with Bennett et al. (2007). an increased body mass index (BMI) was found to be
associated with enhanced FEVi responses to O3 by McDonnell et al. (2010).
The BMI effect is of the same order of magnitude but in the opposite direction of the
age effect whereby FEVi responses diminish with increasing age. Although the
effects of age and BMI are relatively strong, these characteristics account for only a
small amount of the observed variability in individual responses.
Alternatively, Lefohn et al. (2010a) proposed that FEVi responses to O3 exposure
might be described by a cumulative integrated exposure index with a sigmoidal
weighting function similar to the W126 used for predicting vegetation effects (see
Section 9.5). The integrated exposure index is the sum of the hourly average O3
6-14
-------
concentrations times their respective weighing factors. Based on a limited number of
studies, the authors assumed weighting factors ranged from near zero at 50 ppb up to
approximately 1.0 for concentrations at > 125 ppb. The concentrations of 60, 70 and
80 ppb correspond to the assumed weights of 0.14, 0.28, and 0.50, respectively, and
apply only to the case of exposure during moderate exercise
(VE = 20 L/min per m2 BSA). Lefohn et al. (2010a) calculated the cumulative
exposure index for the protocols used by Adams (2006a, 2003a) and Schelegle et al.
(2009). They found statistically significant O3 effects after 4 hours on FEVi at
105 ppb-hour based on Schelegle et al. (2009) and at 235 ppb-hour based on Adams
(2006a, 2003a). Based on this analysis, the authors recommended a 5-hour
accumulation period to protect against O3 effects on lung function.
More recently the McDonnell et al. (2007) model, as well as a variant containing a
response threshold (described in more detail below), was fit to a larger dataset
consisting of the FEVi responses of 741 young healthy adults (104 F, 637 M; mean
age 23.8 yrs) from 23 individual controlled exposure studies conducted in either
Chapel Hill, NC or Davis, CA (McDonnell et al., 2012). Concentrations across
individual studies ranged from 40 ppb to 400 ppb, activity level ranged from rest to
heavy exercise, duration of exposure was from 2 to 7.6 hours, and some studies
provided data during periods of recovery following exposure. The resulting empirical
models can estimate the frequency distribution of individual responses for any
exposure scenario as well as summary measures of the distribution such the mean or
median response and the proportions of individuals with FEVi decrements > 10%,
15%, and 20%. Predictions were found to be close agreement with the experimental
data. The responses of males and females were, on average, approximately equal
when activity level was controlled by normalizing VE to BSA. Thus, any effects of
sex upon FEVi responses to O3 exposure can be accounted for by utilizing VE/BSA.
In this large dataset, the coefficient of BMI was not statistically significantly
different from zero, although its magnitude was similar to that estimated by of the
earlier study (McDonnell et al., 2010). The threshold model fit the experimental data
better than the non-threshold model, particularly at the earliest time points of low
concentration exposures.
Schelegle et al. (2012) also analyzed a large dataset with substantial overlap to that
used by McDonnell et al. (2012). From an initial dataset consisting of the FEVi
responses of 704 young healthy adults (76 F, 628 M; mean age 23.8 yrs) from 21
individual controlled exposure studies conducted in either Chapel Hill, NC or Davis,
CA, their model was fit to the FEVi responses of 220 young healthy adults (51 F,
mean age 22 yrs; 169 M, mean age 24 yrs). Eighty-one of the excluded individuals
appeared to be the result of inherent variability of repeated FEVi measurements in
certain individuals and were present in both FA and O3 exposure protocols.
The resulting model unrealistically overestimated the FEVi responses of 11
individuals that participated in short-duration exposures (2.5 hours) with heavy
exercise (VE = 35 L/min per BSA) and high O3 concentrations (240, 300, and 400
ppb). However, in general, for most exposure scenarios, the authors concluded that
their model coefficients based on 220 individuals reliably predicted the mean FEVi
decrements for the full dataset of 704 individuals.
6-15
-------
Both McDonnell et al. (2012) and Schelegle et al. (2012) developed two
compartment models that considered a dose of onset in response or a threshold of
response. The first compartment in the McDonnell et al. (2012) model considers the
level of oxidant stress in response to O3 exposure to increase over time as a function
of dose rate (CxVE) and decrease by clearance or metabolization over time according
to first order reaction kinetics. In the second compartment of the threshold model,
once oxidant stress reaches some threshold level, the decrement in FEVi increases as
a sigmoid-shaped function of the oxidant stress with age. In the Schelegle et al.
(2012) model, a first compartment acts as a reservoir in which oxidant stress builds
up until the dose of onset at which time it spills over into a second compartment.
The second compartment is identical to the first compartment in McDonnell et al.
(2012) model. The oxidant levels in the second compartment were multiplied by a
responsiveness coefficient to predict FEVi responses for the Schelegle et al. (2012)
model.
Exposures predicted to not reach threshold in the McDonnell et al. (2012) dataset
were those with moderate, near continuous exercise for 1 hour to 60 and 80 ppb O3,
and for 2 hours to 40 ppb; and those at rest for 1 hour to 180 and 240 ppb O3, and for
2 hours to 120 ppb O3. However, there were also exposures above the threshold
having small predicted responses due to the sigmoid shape of the exposure-response
function. Schelegle et al. (2012) reported an average predicted dose of onset in
response was 1,080 ug O3. For a prolonged (6.6 hours) O3 exposure with moderate
quasi continuous exercise (VE = 20 L/min per BSA), this dose of onset would not be
reached until between 4-5 hrs of exposure to 60 ppb or 3-4 hrs of exposure to 80 ppb.
However, 14% of the individuals in the Schelegle et al. (2012) study had a dose of
onset of less than 400 ug O3. More consistent with the threshold in response reported
by McDonnell et al. (2012), this dose of onset (i.e., 400 ug O3) would be reached in
1-2 hrs of exposure to 50-80 ppb O3 with moderate quasi continuous exercise.
Intersubject Variability in Response of Healthy Subjects
Consideration of group mean changes is important in discerning if observed effects
are due to O3 exposure rather than chance alone. Inter-individual variability in
responses is, however, considerable and pertinent to assessing the fraction of the
population that might actually be affected during an O3 exposure. Hackney et al.
(1975) first recognized a wide range in the sensitivity of subjects to O3. The range in
the subjects' ages (29 to 49 years) and smoking status (0 to 50 pack years) in the
Hackney et al. (1975) study are now understood to affect the spirometric and
symptomatic responses to O3. Subsequently, DeLucia and Adams (1977) examined
responses to O3 in six healthy non-smokers and found that two exhibited notably
greater sensitivity to O3. Since that time, numerous studies have documented
considerable variability in responsiveness to O3 even in subjects recruited to assure
homogeneity in factors recognized or presumed to affect responses.
An individual's FEVi response to a 2 hour O3 exposure is generally reproducible
over several months and presumably reflects the intrinsic responsiveness of the
individual to O3 (Hazucha et al., 2003; McDonnell et al., 1985c). The frequency
6-16
-------
distribution of individual FEVi responses following these relatively short exposures
becomes skewed as the group mean response increases, with some individuals
experiencing large reductions in FEVi (Weinmann et al., 1995a; Kulle et al., 1985).
For 2-hour exposures with intermittent exercise causing a predicted average FEVi
decrement of 10%, individual decrements ranged from approximately 0 to 40% in
white males aged 18-36 years (McDonnell et al.. 1997). For an average FEVi
decrement of 13%, Ultman et al. (2004) reported FEVi responses ranging from a 4%
improvement to a 56% decrement in young healthy adults (32 M, 28 F) exposed for
1 hour to 250 ppb O3. One-third of the subjects had FEVi decrements of >15%, and
7% of the subjects had decrements of >40%. The differences in FEVi responses did
not appear to be explained by intersubject differences in the fraction of inhaled O3
retained in the lung (Ultman et al.. 2004).
fj=
30-
f25-
« 20'
<+-
Percent c
3 Ul O (S
n
— i
-
0 ppb
0%
nH
—
60 ppb
16%
1 n
n
70 ppb
—
19%
In n
nn
—
80 ppb
—
29%
Inl In
-10 0 10 20 30 -10 0 10 20 30
FEV, Decrement (%)
-10 0 10 20 30
Note: During each hour of the exposures, subjects were engaged in moderate quasi continuous exercise (40 L/min) for 50 minutes
and rest for 10 minutes. Following the third hour, subjects had an additional 35 minute rest period for lunch. Subjects were
exposed to a triangular O3 concentration profile having the average O3 concentration provided in each panel. As average O3
concentration increased, the distribution of responses became asymmetric with a few individuals exhibiting large FEVi
decrements. The percentage indicated in each panel is the portion of subjects having a FEVi decrement in excess of 10%.
Source: Adapted with permission of American Thoracic Society (Schelegle et al.. 2009).
Figure 6-2 Frequency distributions of FEVi decrements observed by
Schelegle et al. (2009) in young healthy adults (16 F, 15 M)
following 6.6-hour exposures to Os or filtered air.
Consistent with the 1- to 2-hour studies, the distribution of individual responses
following 6.6-hour exposures becomes skewed with increasing O3 exposure
concentration and magnitude of the group mean FEVi response (McDonnell, 1996).
Figure 6-2 illustrates frequency distributions of individual FEVi responses observed
in 31 young healthy adults following 6.6-hour exposures between 0 and 80 ppb O3.
Schelegle et al. (2009) found >10% FEVi decrements in 16, 19, 29, and 42% of
individuals exposed for 6.6 hours to 60, 70, 80, and 87 ppb O3, respectively. Just as
there are differences in mean decrements between studies having similar exposure
scenarios (Figure 6-1 at 80 and 120 ppb), there are differences in the proportion of
individuals affected with >10% FEVi decrements. At 80 ppb O3, the proportion
affected with >10% FEVi decrements was 17% (n = 30) by Adams (2006a)1. 26%
1 Not assessed by Adams (2006a). the proportion was provided in Figure 8-1B of the 2006 O3AQCD (U.S. EPA. 2006b).
6-17
-------
(n = 60) by McDonnell (1996). and 29% (n = 31) by Schelegle et al. (2009).
At 60 ppb O3, the proportion with >10% FEVi decrements was 20% (n = 30) by
Adams (2002V. 3% (n = 30) by Adams (2006aV. 16% (n = 31) by Schelegle et al.
(2009), and 5% (n = 59) by Kim et al. (2011). Based on these studies, the weighted
average proportion of individuals with >10% FEVi decrements is 10% following
exposure to 60 ppb O3 and 25% following exposure to 80 ppb O3. Due to limited
data within the published papers, these proportions were not corrected for responses
to FA exposure during which lung function typically improves in healthy adults. For
example, uncorrected versus O3-induced (i.e., adjusted for response during FA
exposure) proportions of individuals having >10% FEVi decrements in the Adams
(2006a)2 study were, respectively, 3% versus 7% at 60 ppb and 17% versus 23% at
80 ppb. Thus, uncorrected proportions may underestimate the actual fraction of
healthy individuals affected in some studies.
In addition to examining individual responses on a study-by-study basis, the recently
published McDonnell et al. (2012) model can also be utilized to directly calculate the
proportion of individuals expected to experience O3-induced (i.e., adjusted for
response during FA exposure) FEVi decrements of a given magnitude under a
variety of exposure conditions and demographic characteristics. This model was fit to
the data of young healthy adults (104 F, 637 M; 18-36 yrs of age) that participated in
controlled O3 exposure studies conducted in Chapel Hill, NC and Davis, CA.
Figure 6-3 illustrates the proportions of individuals predicted to have greater than
10%, 15%, and 20% O3-induced FEVi decrements a following a 6.6 hour exposure
to O3 with moderate exercise. Consistent with the observed responses of individual
studies cited above, the model predicts that >10% FEVi decrements occur in 9% of
the individuals exposed to 60 ppb and 22% of those exposed to 80 ppb O3.
1 This information is from page 761 of Adams (2002). Adams (2006a. 2002) both provide data for a group of 30 healthy subjects that
were exposed via facemask to 60 ppb (square-wave) O3 for 6.6 hours with moderate exercise (VE = 23 L/min per m2 BSA). These
subjects are described on page 133 of Adams (2006a) and pages 747 and 761 of Adams (2002). The FEVi decrement may be
somewhat increased due to a target VE of 23 L/min per m2 BSA relative to other studies with which it is listed having the target VE
of 20 L/min per m2 BSA. Based on Adams (2003a. b, 2002). similar FEN/! responses are expected between facemask and
chamber exposures.
2 Not assessed by Adams (2006a). uncorrected and O3-induced proportions are from Figures 8-1B and 8-2, respectively, of the
2006 O3AQCD (U.S. EPA. 2006b).
6-18
-------
15 i
o>
§110% FEVi
decrements of 10% (n = 150) (Kim et al.. 2011: Schelegle et al.. 2009: Adams.
2006a. 1998). In an individual with relatively "normal" lung function, with
recognition of the technical and biological variability in measurements, confidence
can be given that within-day changes in FEVi of > 5% are clinically meaningful
(Pellegrino et al.. 2005: ATS. 1991). Here focus is given to individuals with >10%
decrements in FEVi since some individuals in the Schelegle et al. (2009) study
experienced 5-10% FEVi decrements following exposure to FA. A 10% FEVi
decrement is also generally accepted as an abnormal response and a reasonable
criterion for assessing exercise-induced bronchoconstriction (Dryden et al.. 2010:
ATS. 2000a). The data are not available in the published papers to determine the
O3-induced proportion for either the Adams (1998) or Schelegle et al. (2009) studies.
As already stated, however, this uncorrected proportion likely underestimates the
actual proportion of healthy individuals experiencing O3-induced FEVi decrements
6-19
-------
in excess of 10%. Therefore, by considering uncorrected responses and those
individuals having >10% decrements, 10% is an underestimate of the proportion of
healthy individuals that are likely to experience clinically meaningful changes in lung
function following exposure for 6.6 hours to 60 ppb O3 during moderate exercise.
Of the studies conducted at 60 ppb, only Kim et al. (2011) reported FEVi decrements
at 60 ppb to be statistically significant. However, Brown et al. (2008) found those
from Adams (2006a) to be highly statistically significant. Though group mean
decrements are biologically small and generally do not attain statistical significance,
a considerable fraction of exposed individuals experience clinically meaningful
decrements in lung function.
Factors Modifying Responsiveness to Ozone
Physical activity increases VE and therefore the dose of inhaled O3. Consequently,
the intensity of physiological response during and following an acute O3 exposure
will be strongly associated with VE. Apart from inhaled O3 dose and related
environmental factors (e.g., repeated daily exposures), individual-level factors, such
as health status, age, sex, race/ethnicity, race, smoking habit, diet, and socioeconomic
status (SES) have been considered as potential modifiers of a physiologic response to
such exposures.
Responses in Individuals with Pre-existing Disease
Individuals with respiratory disease are of primary concern in evaluating the health
effects of O3 because a given change in function is likely to have more impact on a
person with pre-existing function impairment and reduced reserve.
Possibly due to the age of subjects studied, patients with COPD performing light to
moderate exercise do not generally experience statistically significant pulmonary
function decrements following 1- and 2-hour exposures to < 300 ppb O3 (Kehrl et al.,
1985: Linnetal.. 1983: Linnetal.. 1982a: Solic et al.. 1982). Following a 4-hour
exposure to 240 ppb O3 during exercise, Gong et al. (1997b) found an O3-induced
FEVi decrement of 8% in COPD patients which was not statistically different from
the decrement of 3% in healthy subjects. Demonstrating the need for control
exposures and the presumed effect of exercise, four of the patients in the Gong et al.
(1997b) study had FEVi decrements of >14% following both the FA and O3
exposures. Although the clinical significance is uncertain, small transient decreases
in arterial blood oxygen saturation have also been observed in some of these studies.
Based on studies reviewed in the 1996 and 2006 O3 AQCDs, subjects with asthma
appear to be at least as sensitive to acute effects of O3 as healthy subjects. Horstman
et al. (1995) found the O3-induced FEVi decrement in 17 subjects with mild-to-
moderate asthma to be significantly larger than that in 13 healthy subjects (19%
versus 10%, respectively) exposed to 160 ppb O3 during light exercise (VE of 15
L/min per m2 BSA) for a 7.6-hour exposure. In subjects with asthma, a significant
positive correlation between O3-induced spirometric responses and baseline lung
6-20
-------
function was observed, i.e., responses increased with severity of disease. In the
shorter duration study by Kreit et al. (1989), 9 subjects with asthma also showed a
considerable larger average O3-induced FEVi decrement than 9 healthy controls
(25% versus 16%, respectively) following exposure to 400 ppb O3 for 2 hours with
moderate-heavy exercise (VE = 54 L/min). Alexis et al. (2000) [400 ppb; 2 hours;
exercise, VE = 30 L/min] and Jorres et al. (1996) [250 ppb; 3 hours; exercise, VE = 30
L/min] reported a tendency for slightly greater FEVi decrements in subjects with
asthma than healthy subjects. Several studies reported similar responses between
individuals with asthma and healthy individuals (Scannell et al., 1996; Hiltermann et
al.. 1995: BashaetaL 1994). The lack of differences in the Hiltermann et al. (1995)
[400 ppb; 2 hours; exercise, VE = 20 L/min] and BashaetaL (1994) [200 ppb;
6 hours; exercise, VE = 25 L/min] studies was not surprising, however, given
extremely small sample sizes (5-6 subjects per group) and corresponding lack of
statistical power. Power was not likely problematic for Scannell et al. (1996)
[200 ppb; 4 hours; exercise, VE ~ 44 L/min] with 18 subjects with mild asthma and
81 age-matched healthy controls from companion studies (Balmes et al., 1996; Aris
et al., 1995). Of note, Mudway et al. (2001) reported a tendency for subjects with
asthma to have smaller O3-induced FEVi decrements than healthy subjects (3%
versus 8%, respectively) when exposed to 200 ppb O3 for 2 hours during exercise.
However, the subjects with asthma in Mudway et al. (2001) also tended to be older
than the healthy subjects, which could partially explain their smaller response since
FEVi responses to O3 diminish with age.
In a study published since the 2006 O3 AQCD, Stenfors et al. (2010) exposed
subjects with persistent asthma (n = 13; aged 33 years) receiving chronic inhaled
corticosteroid therapy to 200 ppb O3 for 2 hours with moderate exercise. An average
O3-induced FEVi decrement of 8.4% was observed, whereas, only a 3.0% FEVi
decrement is predicted for similarly exposed age-matched healthy controls
(Mcdonnell et al.. 2007). Vagaggini et al. (2010) exposed subjects with mild-to-
moderate asthma (n = 23; 33 ± 11 years) to 300 ppb O3 for 2 hours with moderate
exercise. Although the group mean O3-induced FEVi decrement was only 4%, eight
subjects were categorized as "responders" with >10% FEVi decrements. Baseline
lung function did not differ between the responders and nonresponders suggesting
that, in contrast to Horstman et al. (1995), O3-induced FEVi responses were not
associated with disease severity.
Lifestage
Children, adolescents, and young adults (<18 years of age) appear, on average, to
have nearly equivalent spirometric responses to O3, but have greater responses than
middle-aged and older adults when similarly exposed to O3 (U.S. EPA, 1996a).
Symptomatic responses to O3 exposure, however, appear to increase with age until
early adulthood and then gradually decrease with increasing age (U.S. EPA, 1996a).
For example, healthy children (n=22; mean age 10 yrs) exposed to FA and 120 ppb
O3 (2.5 hours; heavy intermittent exercise, VE=32-35 L/min per m2 BSA)
experienced similar spirometric responses, but lesser symptoms than similarly
exposed young healthy adults (n=21-22; mean age 22 yrs)(McDonnell et al., 1985a).
6-21
-------
For subjects aged 18-36 years, McDonnell et al. (1999b) reported that symptom
responses from O3 exposure also decrease with increasing age. Diminished
symptomatic responses in children and the elderly might put these groups at
increased risk for continued O3 exposure, i.e., a lack of symptoms may result in their
not avoiding or ceasing exposure. Once lung growth and development reaches the
peak (18-20 years of age in females and early twenties in males), pulmonary
function, which is at its maximum as well, begins to decline progressively with age
as does O3 sensitivity.
A couple analyses of large datasets have assessed the effects of age on
responsiveness to O3 exposure. McDonnell et al. (2007) found O3-induced FEVi
responses to decrease significantly with increasing age in an analysis of data from
studies of healthy, nonsmoking, white males (n = 541), 18-35 years old (mean age
24.1), conducted in Chapel Hill, NC. Using this same dataset, McDonnell et al.
(2010) reported that O3-induced FEVi responses, while decreasing significantly with
age, also increased significantly with BML In a larger dataset of 741 young healthy
adults (104 F, 637 M; mean age 23.8 yrs) from studies conducted in either Chapel
Hill, NC or Davis, CA, McDonnell et al. (2012) did not find a statistically significant
effect of either age or BMI on the FEVi responses. Analysis of the Davis data alone
showed a tendency for increases in O3-induced FEVi responses with increases in
both age and BMI, whereas FEVi responses in the Chapel Hill data decreased with
age and increased with BMI. The authors speculated that the lack of a significant age
effect may be, in part, due to a significant correlation (r = 0.23) between age and
BMI in the 142 subjects from studies conducted at Davis. No correlation (r = 0.03)
between age and BMI was observed in the Chapel Hill data.
In healthy individuals, the fastest rate of decline in O3 responsiveness appears
between the ages of 18 and 35 years (Passannante et al., 1998; Seal et al., 1996),
more so for females then males (Hazucha et al., 2003). During the middle age period
(35-55 years), O3 sensitivity continues to decline, but at a much lower rate. Beyond
this age (>55 years), acute O3 exposure elicits minimal spirometric changes. Whether
the same age-dependent pattern of O3 sensitivity decline also holds for
nonspirometric pulmonary function, airway reactivity or inflammatory endpoints has
not been determined. Although there is considerable evidence that spirometric and
symptomatic responses to O3 exposure decrease with age beyond young adulthood,
this evidence comes from cross-sectional analyses and has not been confirmed by
longitudinal studies of the same individuals.
Sex
Several studies have suggested that physiological differences between sexes may
predispose females to greater O3-induced health effects. In females, lower plasma
and nasal lavage fluid (NLF) levels of uric acid (the most prevalent antioxidant), the
initial defense mechanism of O3 neutralization in airway surface liquid, may be a
contributing factor (Houslev et al.. 1996). Consequently, reduced absorption of O3 in
the upper airways may promote its deeper penetration. Dosimetric measurements
have shown that the absorption distribution of O3 is independent of sex when
6-22
-------
absorption is normalized to anatomical dead space (Bush et al., 1996). Thus, a sex-
related differential removal of O3 by uric acid seems to be minimal. In general, the
physiologic response of young healthy females to O3 exposure appears comparable
to the response of young males (Hazucha et al., 2003). Based on analysis of a large
dataset (104 F, 637 M; 18-36 yrs of age), FEVi responses of males and females to
O3 exposure were, on average, approximately equal when activity level was
controlled by normalizing VE to BSA (McDonnell et al., 2012). Additionally, with
activity level assessed as VE/BSA, Schelegle et al. (2012) observed no significant
difference in the average dose to onset of responses between males and females.
Several studies have investigated the effects of the menstrual cycle on responses to
O3 in healthy young women. In a study of 9 women exposed during exercise to
300 ppb O3 for an hour, Fox et al. (1993) found lung function responses to O3
significantly enhanced during the follicular phase relative to the luteal phase.
However, Weinmann et al. (1995c) found no difference in responses between the
follicular and luteal phases as well as no significant differences between 12 males
and 12 females exposed during exercise to 350 ppb O3 for 2.15 hours. Seal et al.
(1996) also reported no effect of menstrual cycle phase in their analysis of responses
of 150 women (n = 25 per exposure group; 0, 120, 240, 300, and 400 ppb O3). Seal
et al. (1996) conceded that the methods used by Fox et al. (1993) more precisely
defined menstrual cycle phase.
Race/Ethnicity
Only two controlled human exposure studies have assessed differences in lung
function responses between races. Seal et al. (1993) compared lung function
responses of whites (93 M, 94 F) and blacks (undefined ancestry; 92 M, 93 F)
exposed to a range of O3 concentrations (0-400 ppb). The main effects of the sex-
race group and O3 concentration were statistically significant (both at p <0.001),
although the interaction between sex-race group and O3 concentration was not
significant (p = 0.13). These findings indicate some overall difference between the
sex-race groups that is independent of O3 concentration, i.e., the concentration-
response (C-R) curves for the four sex-race groups are parallel. In a multiple
comparison procedure on data collapsed across all O3 concentrations for each sex-
race group, both black men and black women had significantly larger decrements in
FEVi than did white men. The authors noted that the O3 dose per unit of lung tissue
would be greater in blacks and females than whites and males, respectively. It cannot
be ruled out that this difference in tissue dose might have affected responses to O3.
The college students recruited for the Seal et al. (1993) study were noted by the
authors as probably being from better educated and SES advantaged families, thus
reducing the potential influence of these variables on results. In a follow-up analysis,
Seal et al. (1996) reported that, of three SES categories, individuals in the middle
SES category showed greater concentration-dependent decline in percent-predicted
FEVi (4-5% at 400 ppb O3) than low and high SES groups. The authors did not have
an "immediately clear" explanation for this finding.
6-23
-------
More recently, Que et al. (2011) assessed pulmonary responses in blacks of African
American ancestry (22 M, 24 F) and Caucasians (55 M, 28 F) exposed to 220 ppb O3
for 2.25 hours (alternating 15 min periods of rest and brisk treadmill walking).
On average, the black males experienced a 16.8% decrement in FEVi following O3
exposure which was significantly larger than mean FEVi decrements of 6.2, 7.9, and
8.3% in black females and Caucasian males and Caucasian females, respectively.
In the study by Seal et al. (1993). there was potential that the increased FEVi
decrements in blacks relative to whites were due to increased O3 tissue doses since
exercise rates were normalized to BSA. Differences in O3 tissue doses between the
races should not have occurred in the Que et al. (2011) study because exercise rates
were normalized to lung volume (viz., 6-8 times FVC). Thus, the increased mean
FEVi decrement in black males is not likely attributable to systematically larger O3
tissue doses in blacks relative to whites.
Smoking
Smokers are less responsive to O3 for some (but not all) health endpoints than
nonsmokers. Spirometric and plethysmographic pulmonary function decline,
respiratory symptoms, and nonspecific airway hyperreactivity of smokers to O3 were
all weaker than data reported for nonsmokers. However, the time course of
development and recovery of these effects as well their reproducibility in smokers
were not different from nonsmokers (Frampton et al., 1997a). Another similarity
between smokers and nonsmokers is that, the inflammatory response to O3 does not
appear to depend on smoking status or the responsiveness of individuals to changes
in lung function (Torres et al., 1997). Chronic airway inflammation with
desensitization of bronchial nerve endings and an increased production of mucus
may plausibly explain the reduced responses to O3 in smokers relative to nonsmokers
(Frampton et al.. 1997a: Torres etal.. 1997).
Antioxidant supplementation
The first line of defense against oxidative stress is antioxidants-rich ELF which
scavenges free radicals and limits lipid peroxidation. Exposure to O3 depletes the
antioxidant level in nasal ELF probably due to scrubbing of O3 (Mudway et al.,
1999a); however, the concentration and the activity of antioxidant enzymes either in
ELF or plasma do not appear to be related to O3 responsiveness (Samet et al., 2001;
Avissar et al., 2000; Blomberg et al., 1999). Carefully controlled studies of dietary
antioxidant supplementation have demonstrated some protective effects of a-
tocopherol and ascorbate on spirometric lung function from O3 but not on the
intensity of subjective symptoms or inflammatory response including cell
recruitment, activation and a release of mediators (Samet et al.. 2001; Trenga et al..
2001). Dietary antioxidants have also been reported to attenuate O3-induced
bronchial hyperresponsiveness in asthmatics (Trenga et al.. 2001).
Genetic polymorphisms
Some studies (e.g., Corradi et al.. 2002; Bergamaschi et al.. 2001) reviewed in the
2006 O3 AQCD reported that genetic polymorphisms of antioxidant enzymes may
6-24
-------
modulate pulmonary function and inflammatory response to O3 challenge. It was
suggested that healthy carriers of NAD(P)H: quinone oxidoreductase wild type
(NQOlwt) in combination with GSTM1 null were more responsive to O3.
Bergamaschi et al. (2001) reported that subjects having NQOlwt and GSTM1 null
genotypes had increased O3 responsiveness (FEVi decrements and epithelial
permeability), whereas subjects with other combinations of these genotypes were less
affected. A subsequent study from the same laboratory reported a positive association
between O3 responsiveness, as characterized by the level of oxidative stress and
inflammatory mediators (8-isoprostane, LTB4 and TEARS) in exhaled breath
condensate and the NQOlwt and GSTMlnull genotypes (Corradi et al.. 2002).
However, none of the spirometric endpoints (e.g., FEVi) were affected by O3
exposure.
In a controlled exposure of subj ects with mild-to-moderate asthma (n = 23;
33 ± 11 years) to 300 ppb O3 for 2 hours with moderate exercise, Vagaggini et al.
(2010) found that six of the subjects had a NQOlwt and GSTM1 null, but this
genotype was not associated with the changes in lung function or inflammatory
responses to O3. Kim et al. (2011) also recently reported that GSTM1 genotype was
not predictive of FEVi responses to O3 in young healthy adults (32 F, 27 M;
25.0 ± 0.5 year) who were roughly half GSTM1-null and half GSTM1-sufficient.
Sputum neutrophil levels, measured in a subset of the subjects (13 F, 11 M), were
also not significantly associated with GSTM1 genotype.
In a study of healthy volunteers with GSTM1 sufficient (n = 19; 24 ± 3) and GSTM1
null (n = 16; 25 ± 5) genotypes exposed to 400 ppb O3 for 2 hours with exercise,
Alexis et al. (2009) found that inflammatory responses but not lung function
responses to O3 were dependent on genotype. At 4 hours post-O3 exposure, both
GSTM1 genotype groups had significant increases in sputum neutrophils with a
tendency for a greater increase in GSTM1 sufficient than null subjects. At 24 hours
postexposure, sputum neutrophils had returned to baseline levels in the GSTM1
sufficient individuals. In the GSTM1 null subjects, however, sputum neutrophil
levels increased from 4 hours to 24 hours and were significantly greater than both
baseline levels and levels at 24 hours in the GSTM1 sufficient individuals. Since
there was no FA control in the Alexis et al. (2009) study, effects of the exposure
other than O3 itself cannot be ruled out. In general, the findings between studies are
inconsistent.
Body Mass Index
In a retrospective analysis of data from 541 healthy, nonsmoking, white males
between the ages of 18-35 years from 15 studies conducted at the U.S. EPA Human
Studies Facility in Chapel Hill, NC, McDonnell et al. (2010) found that increased
BMI was associated with enhanced FEVi responses to O3. The BMI effect was of
the same order of magnitude but in the opposite direction of the age effect whereby
FEVi responses diminish with increasing age. In a similar retrospective analysis,
Bennett et al. (2007) found enhanced FEVi decrements following O3 exposure with
increasing BMI in a group of 75 healthy, nonsmoking, women (age 24 ± 4 years;
6-25
-------
BMI range 15.7 to 33.4), but not 122 healthy, nonsmoking, men (age 25 ± 4 years;
BMI range 19.1 to 32.9). In the women, greater O3-induced FEVi decrements were
seen in overweight (BMI >25) than in normal weight (BMI from 18.5 to 25), and in
normal weight than in underweight (BMI <18.5) (P trend < 0.022). Together, these
results indicate that higher BMI may be a risk factor for pulmonary effects associated
with O3 exposure.
Repeated Ozone Exposure Effects
The attenuation of responses observed after repeated consecutive O3 exposures in
controlled human exposure studies has also been referred to in the literature as
"adaptation" or "tolerance" (e.g., Linn et al., 1988). In animal toxicology studies,
however, the term tolerance has more classically been used to describe the
phenomenon wherein a prior exposure to a low, nonlethal concentration of O3
provides some protection against death and lung edema at a higher, normally lethal
exposure concentration (see Section 9.3.5 of U.S. EPA, 1986). The term
"attenuation" will be used herein to refer to the reduction in responses to O3
observed with repeated O3 exposures in controlled human exposure studies. Neither
tolerance nor attenuation should be presumed to imply complete protection from the
biological effects of inhaled O3, because continuing injury still occurs despite the
desensitization to some responses.
The attenuation of responses due to ambient O3 exposure was first investigated by
Hackney et al. (1977a); (1976). Experiencing frequent ambient O3 exposures, Los
Angeles residents were compared to groups having less ambient O3 exposure.
Following a controlled laboratory exposure to 370-400 ppb O3 for 2 hours with light
intermittent exercise (2-2.5 times resting VE), the Los Angeles residents exhibited
minimal FEVi responses relative to groups having less ambient O3 exposure.
Subsequently, Linn et al. (1988) examined the seasonal variation in Los Angeles
residents' responses to O3 exposure. A group of 8 responders (3M, 5F) and 9
nonresponders (4M, 5F) were exposed to 180 ppb O3 for 2 hours with heavy
intermittent exercise (VE = 35 L/min per m2 BSA) on four occasions (spring, fall,
winter, and the following spring). In responders, relative to the first spring exposures,
FEVi responses were attenuated in the fall and winter, but returned to similar
decrements the following spring. By comparison, the nonresponders, on average,
showed no FEVi decrements on any of the four occasions. In subjects recruited
regardless of FEVi responsiveness to O3 from the area around Chapel Hill, NC, no
seasonal effect of ambient O3 exposure on FEVi responses following chamber
exposures to O3 has been observed (Hazucha et al., 2003; McDonnell et al., 1985c).
Based on studies reviewed in previous O3 AQCDs, several conclusions can be drawn
about repeated 1- to 2-hour O3 exposures. Repeated exposures to O3 causes
enhanced (i.e., greater decrements) FVC and FEVi responses on the second day of
exposure. The enhanced response appears to depend to some extent on the magnitude
of the initial response (Horvath et al.. 1981). Small responses to the first O3 exposure
are less likely to result in an enhanced response on the second day of O3 exposure
6-26
-------
(Folinsbee et al., 1994). With continued daily exposures (i.e., beyond the second day)
there is a substantial (or even total) attenuation of pulmonary function responses,
typically on the third to fifth days of repeated O3 exposure. This attenuation of
responses is lost in 1 week (Kulle et al., 1982; Linn et al., 1982b) or perhaps 2 weeks
(Horvath et al., 1981) without O3 exposure. In temporal conjunction with pulmonary
function changes, symptoms induced by O3 (e.g., cough, pain on deep inspiration,
and chest discomfort), are also increased on the second exposure day but are
attenuated with repeated O3 exposure thereafter (Folnsbee et al.. 1995: Foxcroft and
Adams. 1986: Linn et al.. 1982b: Folinsbee et al.. 1980). In longer-duration
(4-6.6 hours), lower-concentration studies that do not cause an enhanced second-day
response, the attenuation of response to O3 appears to proceed more rapidly
(Folinsbee et al.. 1994).
Consistent with other investigators, Frank et al. (2001) found FVC and
decrements to be significantly attenuated following four consecutive days of
exposure to O3 (250 ppb, 2 hours). However, the effects of O3 on the small airways
(assessed by a combined index of isovolumetric forced expiratory flow between 25
and 75% of vital capacity [FEF 25-75] and flows at 50% and 75% of FVC) showed a
persistent functional reduction from Day 2 through Day 4. Notably, in contrast to
FVC and FEVi which exhibited a recovery of function between days, there was a
persistent effect of O3 on small airways function such that the baseline function on
Day 2 through Day 4 was depressed relative to Day 1. Frank et al. (2001) also found
neutrophil (PMN) numbers in BAL remained significantly higher following O3 (24
hours after last O3 exposure) compared to FA. Markers from bronchioalveolar lavage
fluid (BALF) following 4 consecutive days of both 2-hour (Devlin et al., 1997) and
4-hour (Jorres et al., 2000: Christian et al., 1998) exposures have indicated ongoing
cellular damage irrespective of the attenuation of some cellular inflammatory
responses of the airways, lung function and symptoms response. These data suggest
that the persistent small airways dysfunction assessed by Frank et al. (2001) is likely
induced by both neurogenic and inflammatory mediators, since the density of
bronchial C-fibers is much lower in the small than large airways.
Summary of Controlled Human Exposure Studies on Lung Function
Responses in humans exposed to O3 concentrations found in the ambient
environment include: decreased inspiratory capacity; mild bronchoconstriction;
rapid, shallow breathing pattern during exercise; and symptoms of cough and pain on
deep inspiration (U.S. EPA, 2006b, 1996a). Discussed in subsequent Section 6.2.2.1
and Section 6.2.3.1, controlled exposure to O3 also results in airway
hyperresponsiveness, pulmonary inflammation, immune system activation, and
epithelial injury (Que et al., 2011; Mudway and Kelly, 2004a). Reflex inhibition of
inspiration results in a decrease in forced vital capacity and, in combination with
mild bronchoconstriction, contributes to a decrease in the FEVi. Healthy young
adults exposed to O3 concentrations > 60 ppb develop statistically significant
reversible, transient decrements in lung function and symptoms of breathing
discomfort if minute ventilation or duration of exposure is increased sufficiently
6-27
-------
(Kim etal.. 2011: McDonnell et al.. 2010: Schelegle et al.. 2009: Brown et al.. 2008:
Adams, 2006a). With repeated O3 exposures over several days, FEVi and symptom
responses become attenuated in both healthy individuals and asthmatics, but this
attenuation of responses is lost after about a week without exposure (Gong et al.,
1997a: Folinsbee et al., 1994: Kulle et al., 1982). In contrast to the attenuation of
FEVi responses, there appear to be persistent O3 effects on small airways function as
well as ongoing cellular damage during repeated exposures.
There is a large degree of intersubject variability in lung function decrements
(McDonnell, 1996). However, these lung function responses tend to be reproducible
within a given individual over a period of several months indicating differences in
the intrinsic responsiveness of individuals (Hazucha et al., 2003: McDonnell et al.,
1985c). In healthy young adults, O3-induced decrements in FEVi do not appear to
depend on sex (Hazucha et al., 2003), body surface area or height (McDonnell et al.,
1997), lung size or baseline FVC (Messineo and Adams, 1990). There is limited
evidence that blacks may experience greater O3-induced decrements in FEVi than
age-matched whites (Que et al., 2011: Seal et al., 1993). Healthy children experience
similar spirometric responses but lesser symptoms from O3 exposure relative to
young adults (McDonnell et al., 1985b). On average, spirometric and symptom
responses to O3 exposure appear to decline with increasing age beyond about
18 years of age (McDonnell et al., 1999b: Seal et al., 1996). There is a tendency for
slightly increased spirometric responses in individuals with mild asthma and allergic
rhinitis relative to healthy young adults (Jorres et al., 1996). Spirometric responses in
subjects with asthma appear to be affected by baseline lung function, i.e., responses
increase with disease severity (Horstman et al., 1995).
Available information on recovery of lung function following O3 exposure indicates
that an initial phase of recovery in healthy individuals proceeds relatively rapidly,
with acute spirometric and symptom responses resolving within about 2 to 4 hours
(Folinsbee and Hazucha, 1989). Small residual lung function effects are almost
completely resolved within 24 hours. One day following O3 exposure, persistent
effects on the small airways assessed by decrements in FEF 25.75 and altered
ventilation distribution have been reported (Frank et al., 2001: Foster et al., 1997).
6.2.1.2 Epidemiology
The O3-induced lung function decrements consistently demonstrated in controlled
human exposure studies (Section 6.2.1.1) provide biological plausibility for the
epidemiologic evidence consistently linking short-term increases in ambient O3
concentration with lung function decrements in diverse populations. In the 1996 and
2006 O3 AQCDs, coherence with controlled human exposure study results was found
not only for epidemiologic associations observed in groups with expected higher
ambient O3 exposures and higher exertion levels, including children attending
summer camps and adults exercising or working outdoors, but also for associations
observed in children and individuals with asthma (U.S. EPA. 2006K 1996a). Recent
6-28
-------
epidemiologic studies focused more on children with asthma rather than groups with
increased outdoor exposures or other healthy populations. Whereas recent studies
contributed less consistent evidence, the cumulative body of evidenc e indicates
decreases in lung function in association with increases in ambient O3 concentration
in children with asthma. Collectively, studies in adults with asthma and individuals
without asthma found both O3-associated decreases and increases in lung function.
Recent studies did provide additional data to assess whether particular lags of O3
exposure were more strongly associated with decrements in lung function; whether
O3 associations were confounded by copollutant exposures; and whether associations
were modified by factors such as corticosteroid (CS) use, genetic polymorphisms,
and elevated BMI.
Populations with Increased Outdoor Exposures
Epidemiologic studies primarily use ambient O3 concentrations to represent
exposure; however, few studies have accounted for time spent outdoors, which has
been shown to influence the relationship between ambient concentrations and
individual exposures to O3 (Section 4.3.3). Epidemiologic studies of individuals
engaged in outdoor recreation, exercise, or work are noteworthy for the likely greater
extent to which ambient O3 concentrations represent ambient O3 exposures. Ambient
O3 concentrations, locations, and time periods for epidemiologic studies of
populations with increased outdoor exposures are presented in Table 6-2. Most of
these studies measured ambient O3 at the site of subjects' outdoor activity and related
lung function changes to the O3 concentrations measured during outdoor activity,
which have contributed to higher O3 personal exposure-ambient concentration
correlations and ratios (Section 4.3.3). Because of improved O3 exposure estimates,
measurement of lung function before and after discrete periods of activity, and
examination of O3 effects during exertion when the dose of O3 reaching the lungs
may be higher due to higher ventilation and inhalation of larger volumes of air,
epidemiologic studies of populations with increased outdoor exposures are more
comparable to controlled human exposure studies. The collective body of
epidemiologic evidence clearly demonstrates decrements in lung function in
association with increases in ambient O3 exposure during outdoor activity
(Figure 6-4 [and Table 6-3]. Figure 6-5 [and Table 6-4]. Figure 6-6 [and Table 6-5].
Expanding upon findings from controlled human exposure studies, these
epidemiologic studies provide strong evidence for respiratory effects in children and
adults related to ambient O3 exposure.
6-29
-------
Table 6-2
Study*
Thurston et al.
(1997)
Berry et al.
(1991)
Spektor and
Lippmann
(1991)
Avol et al.
(1990)
Burnett et al.
(1990)
Higginset al.
(1990)
Raizenne et
al. (1989)
Spektor et al.
(1 988a)
Neas et al.
(1999)
Nickmilderet
al. (2007)
Girardot et al.
(2006)
Korricket al.
(1 998)
Hoppe et al.
(2003)
Spektor et al.
(1 988b)
Selwvn et al.
(1985)
Brunekreef et
al.(1994)
Braun-
Fahrlanderet
al.(1994)
Castilleios et
al.(1995)
Mean and upper percentile O3 concentrations in epidemiologic
studies of lung function in populations with increased outdoor
exposures.
Location
Connecticut
River Valley, CT
Mercer County,
NJ
Fairview Lake,
NJ
Idyllwild, CA
Lake
Couchiching,
Ontario, Canada
San Bernardino,
CA
Lake Erie,
Ontario, Canada
Fairview Lake,
NJ
Philadelphia, PA
Southern
Belgium
Great Smoky
Mountain NP, TN
Mt. Washington,
NH
Munich,
Germany
Tuxedo, NY
Houston, TX
Eastern
Netherlands
Southern
Switzerland
Mexico City,
Mexico
Study Period
June 1991-
1993
July 1988
July-August
1988
June-August
1988
June-July
1983
June-July
1987
June-August
1986
July-August
1984
July-
September
1993
July-August
2002
August-
October 2002
June-August
2003
Summers
1991,1992
Summers
1992-1995
June-August
1985
May-October
1981
June-August
1991
May-October
1989
June 1990-
October 1 991
O3 Averaging Time
1 -h max
1-h max3
1 -h avgb
1 -h avgb
1 -h avgb
1 -h avgb
1 -h avgb
1 -h avgb
1 2-h avga
(9 a.m. - 9 p.m.)
1 -h max
8-h max
Hike-time avg
(2-9 h)d
Hike-time avg
(2-12h)d
30-min max (1-4 p.m.)
Exercise-time avg
(15-55 min)
Exercise-time 15-min
max (4-7 p.m.)
Exercise-time avga
(10-1 45 min)
Exercise-time
30-min avg (1-4 p.m.)
1-h max3
Mean/Median
Concentration
(PPb)
83.6
NR
69
94
59
123
71
53
57.5 (near Camp 1)
55.9 (near Camp 2)
NR
NR
48.1
40
High O3 days: 62.1
Control O3 days:
26.6
NR
47
44.4°
NR
112.3
Upper Percentile
Concentrations (ppb)
Max: 160
Max: 204
Max: 137
Max: 161
Max: 95
Max: 245
Max: 143
Max: 113
Max (near Camp 1): 106
Max (across 6 camps):
24.6-112.8°
Max (across 6 camps):
19.0-81.1°
Max: 74.2
Max: 74
Max (overall): 82
Max: 124
Max: 135
Max: 99.5°
Max: 80°
Max: 365
6-30
-------
Study*
Hoeketal.
(1 993)
Hoppe et al.
(1995)
Chan and Wu
(2005)
Braueret al.
(1996)
Romieu et al.
(1998b)
Thaller etal.
(2008)
Location
Wageningen,
Netherlands
Munich,
Germany
Taichung City,
Taiwan
British Columbia,
Canada
Mexico City,
Mexico
Galveston, TX
Study Period
May-July 1989
April-
September
1993
November-
December
2001
June-August
1993
March-August
1996
Summers
2002-2004
O3 Averaging Time
1-h max3
30-min max (1-4 p.m.)
8-h avg (9 a.m.-
5 p.m.)
1 -h max
1 -h max3
Work-shift avg (mean
9h)3
1 -h max3
Mean/Median
Concentration
(PPb)
NR
High O3 days: 64
Control O3 days: 32
35.6
52.6
40
67.3
35 (median)
Upper Percentile
Concentrations (ppb)
Max: 105°
Max (overall): 77
Max: 65.1
95.5
Max: 84
95th: 105.8
Max: 118
* Note: Studies presented in order of first appearance in the text of this section.
NR = not reported.
3Some or all measurements obtained from monitors located off site of outdoor activity.
b1-h avg preceding lung function measurement, as reported in the pooled analysis by Kinnev etal. (1996).
""Concentrations converted from ug/m3to ppb using the conversion factor of 0.51 assuming standard temperature (25°C) and
pressure (1 atm).
dlndividual-level estimates calculated from concentrations measured in different segments of hiking trail.
Children Attending Summer Camps
Studies of children attending summer camps, most of which were discussed in the
1996 O3 AQCD, have provided important evidence of the effect of ambient O3
exposure on respiratory effects in young, healthy children. In addition to the
improved exposure assessment as described above, these studies were noted for their
daily assessment of lung function by trained staff over 1 - to 2-week periods in the
mornings and late afternoons before and after hours of outdoor activity (Thurston et
al.. 1997: Berry et al.. 1991: Spektor and Lippmann. 1991: Avol et al.. 1990: Burnett
etal.. 1990: Higgins et al.. 1990: Raizenne et al.. 1989: Spektor et al.. 1988a).
In groups mostly comprising healthy children (ages 7-17 years), decrements in FEVi
were associated consistently with increases in ambient O3 concentration averaged
over the 1-12 hours preceding lung function measurement (Figure 6-4 [and
Table 6-31). Kinnev et al. (1996) corroborated this association in a re-analysis
combining 5,367 lung function measurements collected from 616 healthy children
from six studies (Spektor and Lippmann, 1991: Avol et al., 1990: Burnett et al.,
1990: Higgins et aL 1990: Raizenne et al.. 1989: Spektor et al.. 1988a). Based on
uniform statistical methods, a -20 ml (95% CI: -25, -14) change in afternoon FEVi
was estimated for a 40-ppb increase in O3 concentration averaged over the 1 hour
before lung function measurement (Kinnev et al., 1996) (all effect estimates are
standardized to increments specific to the O3 averaging time as detailed in
Section 2.5). All of the studies in the pooled analysis were conducted during summer
months but were diverse in locations examined (i.e., Northeast U.S., Canada,
6-31
-------
California), range in ambient concentrations of O3 (presented within Table 6-2) and
other pollutants measured, and magnitudes of association observed. Study-specific
effect estimates ranged between a 0.76 and 48 mL decrease or between a 0.3% and
2.2% decrease in study mean FEVi per 40-ppb increase in 1-h avg O3.
Among camp studies (including the pooled analysis, plus others), associations for
peak expiratory flow (PEF) were more variable than were those for FEVi, as
indicated by the wider range in effect estimates and wider 95% CIs (Figure 6-4 [and
Table 6-31). Nonetheless, in most cases, increases in ambient O3 concentration were
associated with decreases in PEF. The largest O3-associated decrease in PEF (mean
2.8% decline per 40-ppb increase in 1-h max O3) was found in a group of campers
with asthma, in whom an increase in ambient O3 concentration also was associated
with increases in chest symptoms and bronchodilator use (Thurston et al., 1997).
For both FEVi and PEF, the magnitude of association was not related to the study
mean ambient 1-h avg or max O3 concentration. With exclusion of results from
Spektor and Lippmann (1991). larger O3-associated FEVi decrements were found in
populations with lower mean FEVi • No such trend was found with mean PEF.
Sufficient data were not available to assess whether the temporal variability in O3
concentrations, activity levels of subjects, or associations with other pollutants
contributed to between-study heterogeneity in O3 effect estimates.
6-32
-------
Study
FEV^ (mil
Spektoretal. (1988a)
Spektorand Lippmann
(1991)
Burnett et al. (1990)
Raizenne et al. (1989)
Avoletal. (1990)
Higginsetal. (1990)
Kinneyetal. (1996)
Berry etal. (1991)
PEF (ml/secl
Spektoretal. (1988a)
Burnett etal. (1990)
Raizenne et al. (1989)
Avoletal. (1990)
Higginsetal. (1990)
Kinneyetal. (1996)
Thurston et al. (1997)
Neaset al. (1999)
Berry etal. (1991)
Population
Campers without asthma
Campers without asthma
Campers without asthma
Campers without asthma
Campers without asthma
Campers without asthma
Pooled estimate
Campers without asthma
Campers without asthma
Campers without asthma
Campers without asthma
Campers without asthma
Campers without asthma
Pooled estimate
Campers with asthma
Campers without asthma
Campers without asthma
-160 -120 -80 -40 0 40 80
Change in FEV1 (ml) per unit increase in O3 (95% Cl)
-160 -120 -80 -40 0 40 80
Change in PEF (ml/sec) per unit increase in O3 (95% Cl)
Note: Results generally are presented in order of increasing mean ambient O3 concentration. Effect estimates are from single-
pollutant models and are standardized to a 40-ppb increase for 1-h avg or 1-h max O3 concentration and a 30-ppb increase for
12-h avg O3 concentration.
Figure 6-4 Changes in FEVi (ml_) or PEF (mL/sec) in association with ambient
O3 concentrations among children attending summer camp.
6-33
-------
Table 6-3 Changes in FEVi or PEF in association with ambient O3
concentrations among children attending summer camp for studies
presented in Figure 6-4.
Study* Location
FEVi
Spektoretal.
Spektor and _ . . . . ...
Lippmann(1991) Falrvlew Lake' NJ
Burnett et al. Lake Couchiching,
(1990) Ontario, Canada
Raizenne et al. Lake Erie, Ontario,
(1989) Canada
Av°letal.(1990) Pine Springs, CA
(H1l9^setaL San Bernardino, CA
Kinnev et al. Pooled analysis of
(1996) preceding 6 studies
PEF
Spektoretal.
Burnett et al. Lake Couchiching,
(1990) Ontario, Canada
Raizenne et al. Lake Erie, Ontario,
(1989) Canada
Av°letal.(1990) Pine Springs, CA
,H;^"SetaL San Bernardino, CA
( laaU)
Kinnev et al. Pooled analysis of
(1 996) preceding 6 studies
^asQ.etaL Philadelphia, PA
( I yyy )
Rprrv pt a I
Population, Mean FEVi
(mL) or PEF (mL/sec)
91 campers without asthma
ages 8-1 Syr, 2,140
46 campers without asthma
ages 8-1 4 yr, 2,390
29 campers without asthma
ages 7-1 Syr, 2,410
1 1 2 campers without asthma
mean age 11.6yr, 2,340
295 campers without asthma
ages 8-1 7 yr, 2,190
43 campers without asthma
ages 7-1 Syr, 2,060
61 6 campers without asthma
ages 7-1 7 yr, 2,300
1 4 campers without asthma
age <14yr, NA
91 campers without asthma
ages 8-1 Syr, 4,360
29 campers without asthma
ages 7-1 Syr, 5,480
1 1 2 campers without asthma
mean age 11.6yr, 5,510
295 campers without asthma
ages 8-1 7 yr, 4,520
43 campers without asthma
ages 7-1 Syr, 5,070
61 6 campers without asthma
ages 7-1 7 yr, 4,222
166 campers with asthma
ages 7-1 Syr, 5,333
156 campers without asthma
ages 6-11 yr, 4,717
1 4 campers without asthma
age <14yr, NA
Standardized
Percent Change
(95% Cl)a
-0.93 (-1 .5, -0.35)"
-2.2 (-3.0, -1.3)"
-0.32 (-1.7, 1.1)"
-0.50 (-0.83, -0.16)"
-0.58 (-1.0, -0.12)"
-1 .6 (-2.4, -0.87)"
-0.87 (-1.1, -0.63)
NA
-1.8 (-3.3, -0.40)"
-1.9 (-3.8, -0.05"
-0.07 (-0.56, 0.41)"
1.9(0.71,3.1)"
-0.87 (-2. 1,0.34)"
0.1 6 (-0.45, 0.77)"
-2.8 (-4.9, -0.59)
-0.58 (-1.5, 0.33)
NA
'Includes studies from Figure 6-4.
NA = Data not available.
aAII results are standardized to a 40-ppb increase in 1 -h avg or 1 -h max O3, except that from Neas et al.
standardized to a 30-ppb increase in 12-h avg (9 a.m. -9 p.m.) O3.
"Effect estimates based on results reported in the pooled analysis by Kinnev et al. (1996).
Standardized Effect
Estimate (95% Clf
(mL)
-20.0 (-32.5, -7.5)"
-51 .6 (-72.8, -30.4)"
-7.6 (-42.1, 26.9)"
-11. 6 (-19.4, -3.8)"
-12.8 (-23.0, -2.6)"
-33.6 (-49.3, -17.9)"
-20.0 (-25.5, -14.5)"
32.8 (6.9, 58.7)
(mL/sec)
-80.0 (-142.7, -17.3)"
-106.4 (-209.9, -2.9)"
-4.0 (-30.7, 22.7)"
86.8(31.9,142)"
-44.0 (-105, 17.2)"
6.8 (-19. 1,32.7)"
-146.7 (-261 .7, -31.7)
-27.5 (-70.8, 15.8)
-40.4 (-132.1, 51.3)
(1999). which is
6-34
-------
Similar to controlled human exposure studies, some camp studies found
interindividual variability in the magnitude of O3-associated changes in lung
function. Based on separate regression analyses of serial measurements from
individual subjects, increases in ambient O3 concentration were associated with a
wide range of changes in lung function across subjects (Berry et al., 1991; Higgins et
al.. 1990: Spektor et al.. 1988a). For example, among children attending camp in
Fairview Lake, NJ, 36% of subjects had statistically significant O3-associated
decreases in FEVi, and the 90th percentile of response was a 6.3% decrease in FEVi
per a 40-pbb increase in 1-h avg O3 (Spektor et al.. 1988a).
In contrast with previous studies, a recent study of children attending six different
summer camps in Belgium did not find an association between ambient O3
concentration and lung function (Nickmilder et al., 2007). This study examined
similar ambient O3 concentrations as did previous studies (Table 6-2) but used a less
rigorous methodology. Lung function was measured only once in each subject, and
mean lung function was compared among camps. Children at camps with higher
daily 1-h max or 8-h max O3 concentrations did not consistently have larger
decreases in mean intraday FEVi or FEVi/FVC (Nickmilder et al., 2007).
Populations Exercising Outdoors
As discussed in the 1996 and 2006 O3 AQCDs, epidemiologic studies of adults
exercising outdoors have provided evidence for lung function decrements in healthy
adults associated with increases in ambient O3 exposure during exercise with
durations (10 min to 12 hours) and intensities (heart rates 121-190 beats per min) in
the range of those examined in controlled human exposure studies (Table 6-1).
Associations were found consistently in studies of adults exercising outdoor for up to
2 hours, which similar to the camp studies, measured lung function before and after
exercise by trained staff on multiple occasions. Collectively, studies of exercising
adults found FEVi decrements of 1.3 to 1.5% per unit increase in O3: (Figure 6-5
[and Table 6-41). The magnitude of association did not appear to be related to study
mean ambient O3 concentrations (Table 6-2), exercise duration, or the mean heart
rate measured during exercise (Figure 6-5 [and Table 6-41). Increases in ambient O3
concentration generally were associated with decreases in lung function in the
smaller body of studies of children exercising outdoors (Table 6-4).
1 Effect estimates were standardized to a 40-ppb increase in O3 averaged over 15 min to 1 h and a 30-ppb increase for O3 averaged
over 2 to 12 hours.
6-35
-------
Study
Population
Exercise Mean heart
duration rate(bpm)-1
Brunekreefetal. (1994) Adults exercising 10min-1h 161,176
Spektoretal. (1988b) Adults exercising 15-55min 162,145
Hoppeetal. (2003) Adults exercising 2h NR
Girardotetal. (2006) Adults hiking 2-9h max:121
Korricketal. (1998) Adultshiking 2-12h max: 122
-4
-3
-2
-1
Percentchangein FEV1 per unit increase
in O3 (95% Cl)
Note: Studies generally are presented in order of increasing duration of outdoor exercise. Data for mean heart rate refer to the
maximum or mean measured during exercise or in different groups or conditions as described in Table 6-4.
abpm = beats per minute. NR = Not reported. Effect estimates are from single-pollutant models and are standardized to a 40-ppb
increase for O3 concentrations averaged over 15 minutes to 1 hour and a 30-ppb increase for O3 concentrations averaged over 2
to 12 hours.
Figure 6-5 Percent change in FEVi in association with ambient O3
concentrations among adults exercising outdoors.
6-36
-------
Table 6-4 Percent change in FEVi in association with ambient O3
concentrations among adults exercising outdoors for studies
presented in Figure 6-5, and among children exercising outdoors.
Study*
Location
Population
Exercise
Duration, Mean
Heart Rate
03
Averaging
Time Parameter
Standardized
Percent Change
(95% Cl)a
Studies of adults
Brunekreef
etal.
(1994)
Spektor et
al.(1988b)
Hoppe et
al. (2003)
Girardot et
al. (2006)
Korrick et
al.(1998)
Selwvn et
al. (1985)
Eastern
Netherlands
Tuxedo, NY
Munich,
Germany
Great Smoky
Mt, TN
Mt.
Washington,
NH
Houston, TX
29 adults
exercising,
ages 1 8-37 yr
30 adults
exercising,
ages 21-44yr
43 adults and
children
exercising,
ages 1 3-38 yr
354 adult day
hikers,
ages 1 8-82 yr
530 adult day
hikers,
ages 1 8-64 yr
24 adults
exercising,
ages 29-47 yr
10 min -2.4h,
HR: 161 bpm
(training), 176 bpm
(races)
15-55 min,
HR:162bpm ifVE
>100 L, 145 bpm if
VE 60-100 L
2h, HR:NR
1.8-9 h, max
HR:121 bpm
2-12 h, max HR:
1 22 bpm
Duration: NR,
max HR:
179 bpm in males,
183 bpm in
females
Exercise FEN/!
duration PEF
30-min avg FEV,
30-min max FEVi
(1-4 p.m.) PEF
FEV,
Hike duration
FEV,
Hike duration
PEF
15-minmax FEN/!
-1 .3 (-2.2, -0.37)
-2.5 (-3.8, -1 .2)
-1 .3 (-2.0, -0.64)
-1.3 (-2.6, 0.10)
-2.8 (-5.9, 0.31)
0.72 (-0.46, 1 .90)
3.5 (-0.1 1,7.2)
-1 .5 (-2.8, -0.24)
-0.54 (-4.0, 2.9)
-16 ml (-28.8,
-3.2)b
Studies of children not included in Figure 6-4.
Braun-
Fahrlander
etal.
(1994)
Castilleios
etal.
(1995)
Hoeketal.
(1 993)
Southern
Switzerland
Mexico City,
Mexico
Wageningen,
Netherlands
'Includes studies from Figure 6-5,
128 children
exercising,
ages 9-11 yr
40 children
exercising,
ages 7-11 yr
65 children
exercising,
ages 7-1 2 yr
plus others.
10 min, max
HR: 180 bpm
2 periods, each
with 15 min
exercise and
15 min rest, max
HR:<190bpm
25 min-1.5h,
HR:NR
30-min avg PEF
1-h avg over
combined FEV
exercise-rest 1
period
1-h avg during
exercise
-3.8 (-6.7, -0.96)
-0.48 (-0.72, -0.24)
1 .9 (0.83, 3.0)
HR = heart rate, bpm = beats per minute, VE = minute ventilation, NR = Not reported.
"Effect estimates are standardized to a 40-ppb increase for O3 concentrations averaged over 15 minutes to 1 hour and a 30-ppb
increase for O3 concentrations averaged over 2 to 12 hours.
bResults not included in the figure because data were not available to calculate percent change in lung function.
6-37
-------
Compared with the studies of individuals exercising outdoors described above,
studies of day-hikers assessed lung function only on one day per subject but
examined longer periods of outdoor activity and included much larger sample sizes.
Studies of adult day-hikers had similar design but produced contrasting results
(Girardot et al., 2006; Korrick et al., 1998). Among 530 hikers on Mt. Washington,
NH, Korrick et al. (1998) reported posthike declines in FEVi and FVC of 1.5% and
1.3%, respectively, per a 30-ppb increase in 2- to 12-h avg O3. Associations with
FEVi/FVC, FEF25-75%, and PEF were weaker. In contrast, among 354 hikers on Great
Smoky Mt, TN, Girardot et al. (2006) found that higher O3 concentrations were
associated with posthike increases in many of the same lung function indices
(Figure 6-5 [and Table 6-41). These studies were similar in the examination of a
mostly white, healthy population and of changes in lung function associated with
ambient O3 concentrations measured on site during multihour (2-12 hours) periods of
outdoor exercise. Mean O3 concentrations were similar as were the population mean
and variability in lung function. However, Girardot et al. (2006) differed from
Korrick et al. (1998) in several aspects, including a shorter hike time (mean: 5 versus
8 hours), older age of subjects (mean: 43 versus 35 yr), and measurement of lung
function by a larger number of less well-trained technicians. The impact of these
differences on the heterogeneity in results between the studies was not examined.
Similar to the camp studies, some studies of outdoor exercise examined and found
interindividual variability in the magnitude of O3-associated decreases in lung
function. In Korrick et al. (1998), although a 30-ppb increase in 2- to 12-h avg
ambient O3 concentration was associated with a group mean change in FEF25-75% of-
0.81% (95% CI: -4.9, 3.3), some individuals experienced a >10% decline. The odds
of >10% decline in FEF25-75% increased with increasing ambient O3 concentration
(OR: 2.3 [95% CI: 1.2, 6.7] per 30-ppb increase in 2- to 12-h avg O3). Likewise,
Hoppe et al. (2003) found that compared with days with 30-min max (1-4 p.m.)
ambient O3 concentrations <40 ppb, on days with O3 >50 ppb, 14% of athletes had at
least a 10% decrease in lung function or 20% increase in airway resistance.
Outdoor Workers
Consistent findings in outdoor workers add to the evidence that short-term increases
in ambient O3 exposure can decrease lung function in healthy adults (Figure 6-6 [and
Table 6-51). Except for Hoppe et al. (1995), studies used central site ambient O3
concentrations. However, in outdoor workers, ambient concentrations have been
more highly correlated with and similar in magnitude to personal exposures
(Section 4.3.3) likely because workers spend long periods of time outdoors (6-14
hours across studies) and the O3 averaging times examined correspond to periods of
outdoor work. For example, in a subset of berry pickers, the correlation and ratio of
personal to ambient 24-h avg O3 concentrations (15 km from work site) were 0.64
and 0.96, respectively (Brauer and Brook. 1997). The 6-h avg personal-ambient ratio
in a population of shoe cleaners in Mexico City, Mexico, was 0.56 (O'Neill et al..
2003). Many studies of outdoor workers found that in addition to same-day
concentrations, O3 concentrations lagged 1 or 2 days (Chan and Wu. 2005: Brauer et
6-38
-------
al., 1996) or averaged over 2 days (Romieu et al., 1998b) were associated with equal
or larger decrements in lung function (Figure 6-6 [and Table 6-51).
Similar to other populations with increased outdoor exposure, most of the
magnitudes of O3-associated lung function decrements in outdoor workers were
small, i.e., <1% to 3.4% per unit increase in O3 concentration1. The magnitude of
decrease was not found to depend strongly on duration of outdoor work or ambient
O3 concentration. The largest decrease (6.4% per 40-ppb increase in 1-h max O3)
was observed in berry pickers in British Columbia who were examined during a
period of relatively low ambient O3 concentrations (work shift mean: 26.0 ppb [SD:
11.8]) but had long daily periods of outdoor work (8-14 hours) (Brauer et al., 1996)
(Figure 6-6 [and Table 6-51). However, a much smaller O3-associated decrease in
FEVi was found in shoe cleaners in Mexico City who were examined during a
period of higher O3 concentrations (work shift mean: 67.3 ppb [SD: 24]) but had a
period of outdoor work that was as long as that of the berry pickers. The smallest
magnitude of decrease (-0.4% [95% CI: -0.8, 0] in afternoon FEVi/FVC per 40-ppb
increase in 1-h max O3) was observed in lifeguards in Galveston, TX (Thaller et al.,
2008) whose outdoor work periods were shorter than those of the berry pickers but
characterized by a similar range of ambient O3 concentrations. Not all studies
provided information on ventilation rate or pulse rate, thus it was not possible to
ascertain whether differences in the magnitude of O3-associated lung function
decrement across studies were related to differences in the level of exertion of among
the various groups of workers.
1 Effect estimates were standardized to a 40-ppb increase for O3 averaged over 30 minutes to 1 hour and a 30-ppb increase for O3
averaged over 8 hours or 12 hours.
6-39
-------
Study Population
Thalleretal. (2008) Lifeguards
Parameter O3 Lag Subgroup
FVC 0
FEV^FVC
Braueretal. (1996) Berrypickers FEV, 0
1
Hoppeetal. (1995) Forestry workers FEV,
Romieuetal. (1998) ShoeCleaners FEV,
0
0 Placebo
Antioxidant supplement
0-1 avg Placebo
Antioxidant supplement
-8 -7 -6 -5 -4 -3-2-101 2
Percentchangein lung function perunit
increase in O3 (95% Cl)
Note: Studies generally are presented in order of increasing mean ambient O3 concentration. Effect estimates are from single-
pollutant models and are standardized to a 40-ppb increase for 30-min, 1 -h avg, or 1 -h max O3 concentrations.
Figure 6-6 Percent change in lung function in association with ambient O3
concentrations among outdoor workers.
6-40
-------
Table 6-5 Percent change in FEVi or FEVi/FVC in association with ambient
O3 concentrations among outdoor workers for studies presented in
Figure 6-6.
Study*
Thaller
(2008)
Brauer
etal.
(1996)
Hoppe et
al.
(1995)
Romieu
etal.
(1998b)
Chan
and Wu
(2005)"
'Includes
Location
Galveston,
TX
British
Columbia,
Canada
Munich,
Germany
Mexico City,
Mexico
Taichung
City, Taiwan
Outdoor
Work
Population Parameter Duration
FVC
142 lifeguards, ^ 0 u
ages 1 6-27 yr °'u h
FEV^FVC
58 berry
pickers, ages FEN/! 8-1 4 h
10-69yr
41 forestry .. .
workers, ages FEV, ^
47 male shoe
cleaners, mean __.. Mean (SD):
(SD) age: 38.9 l"tVl 9(1)h
(101 vr
43 mail carriers. M^htti^=
Mean (SD) age: Nlgph"'me 8h
39 (8) yr Ktl~
Os Averaging
Time
1 -h max
1 2-h avg
(7 a.m. -7 p.m.)
1-h max
1 2-h avg
(7 a.m. -7 p.m.)
1 -h max
30-min max
(1 -4p.m.)
1 -h avg before
lung function
measurement
1-h max
8-h avg
(9 a.m. - 5 p.m.)
03
Lag Subgroup
0.
0
1
0
Placebo
0
Antioxidant
0-1 Placebo
avg Antioxidant
0
1
0
1
Standardized
Percent
Change
(95% Cl)a
0.24 (-0.28, 0.72)
0.1 5 (-0.60, 0.90)
-0.40 (-0.80, 0)
-0.60 (-1 .2, 0)
-5.4 (-6.5, -4.3)
-6.4 (-8.0, -4.7)
-1.4 (-3.0, 0.16)
-2.1 (-3.3, -0.85)
-0.52 (-2.0, 0.97)
-3.4 (-6.0, -0.78)
-1.2 (-4.2, 1.8)
-1 .3 (-1 .7, -0.92)
-1.4 (-1.7, -1.2)
-1.6 (-2.2, -1.1)
-1.9 (-2.5, -1.3)
results from Fiaure 6-6, plus others.
"Effect estimates are standardized to a 40-ppb increase for 30-min, 1 -h avg, or 1 -h max O3 and a 30-ppb increase for 8-h avg or
12-h avg O3.
bPEF results not included in figure.
Associations at Lower Ozone Concentrations
In some studies of populations with increased outdoor exposures, O3-associated lung
function decrements were observed when maximum or average ambient O3
concentrations over 30 minutes to 12 hours did not exceed 80 ppb (Chan and Wu,
2005: Korrick et al.. 1998: Hoppe etal.. 1995: Braun-Fahrlander et al.. 1994)
(presented within Table 6-2). Korrick et al. (1998) found lung function decrements in
association with higher hike-time average (2-12 hours) O3 concentrations in the
range 40-74 ppb but not <40 ppb. Several other studies that included higher
maximum ambient O3 concentrations restricted analyses to observations with 10-min
to 1-h avg O3 concentrations <80 ppb (Table 6-6). Higgins et al. (1990) found that
O3-associated lung function decrements in children attending camp were limited
largely to 1-h avg ambient concentrations >120 ppb; however, many other studies
found associations in the lower range of O3 concentrations (Table 6-6). Among
adults exercising outdoors, Spektor et al. (1988b) found that for most lung function
6-41
-------
parameters, effect estimates in analyses restricted to 30-min max ambient O3
concentrations <80 ppb were similar to those obtained for the full range of O3
concentrations (Table 6-6). In a study of children attending summer camp, similar
effects were estimated for the full range of 1-h avg O3 concentrations and those
<60 ppb (Spektor et al, 1988a). Brunekreef et al. (1994) found increases in ambient
O3 concentration (10-min to 1-hour) during outdoor exercise to be associated with
decreases in FEVi in analyses restricted to concentrations <61 (Table 6-6) and
<51 ppb (quantitative results not reported). Whereas Brunekreef et al. (1994) found
that effect estimates were near zero with O3 concentrations <41 ppb, Brauer et al.
(1996) found that associations persisted with 1-h max O3 concentrations <40 ppb
(quantitative results not provided).
Table 6-6 Associations between ambient O3 concentration and FEV1
decrements in different ranges of ambient O3 concentrations.
Study
Brunekreef et al.
(1 994)
Spektor et al.
(1 988a)
Spektor et al.
(1988b)
Korricket al.
(1 998)
Higginset al.
(1990)
Location
Eastern
Netherlands
Fairview Lake,
NJ
Tuxedo, NY
Mt.
Washington,
NH
San
Bernardino,
CA
Population
29 adults exercising,
ages 18-37 yr
91 children without
asthma at camp,
ages 8-1 Syr
30 adults exercising,
ages 21-44yr
530 adult day hikers,
ages 18-64yr
43 children without
asthma at camp,
ages 7-13 yr
Os Averaging
Time
10-min to 2.4-h avg
during exercise
1-h avg before
afternoon FEN/!
measurement
30-min avg during
exercise
2-12 h avg during
hike
1-h avg around time
of FEV,
measurement
03
Concentration
Range
Full range
O3 <61 ppb
Full range
O3 <80 ppb
O3 <60 ppb
Full range
O3 <80 ppb
Full range
O3 40-74 ppb
>120 ppb
<120 ppb
Standardized
Percent
Change
(95% Clf
-1 .3 (-2.2, -0.37)
-2.1 (-4.5, 0.32)
-2.7 (-3.3, -2.0)
-1 .4 (-2.5, -0.34)
-2.2 (-3.7, -0.80)
-1 .3 (-2.0, -0.64)
-1 .3 (-2.4, -0.08)
-1.5 (-2.8, -0.24)
-2.6 (-4.9, -0.32)
-1.4 (-2.8, 0.03)
0.35 (-1.3, 2.0)
"Results are presented in order of maximum O3 concentration included in models. Effect estimates are standardized to a 40-ppb
increase for O3 concentrations averaged over 10 min to 1 h and a 30-ppb increase for O3 concentrations averaged over 2 to 12 h.
Children with Asthma
Increases in ambient O3 concentration have been associated with lung function
decrements in children with asthma in epidemiologic studies conducted across
diverse geographical locations and a range of ambient O3 concentrations (Table 6-7).
Whereas most studies of populations with increased outdoor exposures monitored O3
concentrations at the site of subjects' outdoor activities and used trained staff to
measure lung function, studies of children with asthma relied more on O3 measured
at central monitoring sites and lung function measured by subjects. However, these
methods for exposure and outcome assessment in studies of children with asthma
likely are sources of nondifferential measurement error. Further, compared with the
6-42
-------
camp studies, studies of children with asthma have provided an understanding of the
changes in lung function associated with patterns of outdoor activity and ambient O3
exposure that likely better represent those of children in the general population.
These studies also have provided more information on potential at-risk populations
for O3-associated respiratory effects and on potential confounding by copollutant
exposure or meteorology.
Table 6-7
Study*
Jalaludin et al.
(2000)
Lewis et al.
(2005)
Just et al. (2002)
Hoppe et al.
(2003)
Thurston et al.
(1997)
Romieu et al.
(2006): (2004b:
2002)
Romieu et al.
(1 997)
Romieu et al.
(1996)
O'Connor et al.
(2008)
Mortimer et al.
(2002)
Mortimer et al.
(2000)
Gielen et al.
(1997)
Liu et al.
(2009a). Dales et
al. (2009)
Mean and upper percentile concentrations of O3 in epidemiologic
studies of lung function in children with asthma.
Location
Sydney, Australia
Detroit, Ml
Paris, France
Munich, Germany
CT River Valley, CT
Mexico City, Mexico
Southern Mexico City,
Mexico
Northern Mexico City,
Mexico
Boston, MA; Bronx,
Manhattan NY; Chicago, IL;
Dallas, TX, Seattle, WA;
Tucson, AZ (ICAS)
Bronx, East Harlem, NY;
Baltimore, MD; Washington,
DC; Detroit, Ml,
Cleveland, OH; Chicago, IL;
St. Louis, MO (NCICAS)
Amsterdam, Netherlands
Windsor, ON, Canada
Study Period
Feb-Dec1994
Feb2001-
May 2002
April-June
1996
Summers
1 992-1 995
June 1991-
1993
Oct 1 998-
Apr 2000
Apr-July 1991;
Nov 1991-
Feb 1992
Apr-July 1991;
Nov 1991-
Feb1992
Aug 1998-
July 2001
June-Aug 1993
Apr-July 1995
Oct-Dec2005
03
Averaging
Time
15-h avg
(6a.m.-
9 p.m.)
1-h max
24-h avg
8-h max
24-h avg
30-min max
(1 p.m.-
4 p.m.)
1-h max
8-h max
1-h max
1-h max
1-h max
24-h avg
8-h avg
(10a.m.-
6 p.m.)
8-h max
24-h avg
1-h max
Mean/Median
Concentration
(PPb)
12
26
27.6, 26.5a
40.4, 41. 4a
30.0b
High O3 days:
66.9°
Control O3 days:
32.5°
83.6°
69
102
196
190
NR
48
34. 2b
13.0
27.2
Upper Percentile
Concentrations
(PPb)
Max: 43
91
Overall max: 66.3a
Overall max: 92.0a
Max:61.7b
Max: 91
high O3 days0
Max: 39
Control O3 days0
Max: 160°
Max: 184
Max: 309
Max: 390
Max: 370
NR
NR
Max: 56.5b
95th: 26.5
75th: 32.8
6-43
-------
Study* Location
Rabinovitch et al.
Barraza-Villarreal
Wiwatanadate
fSkultivakorn Chian9 Mai< Thailand
(2010)
Delfino et al.
(2004) Alpme' CA
Hernandez-
Cadena et al. Mexico Citv, Mexico
(2009)
Study Period
Nov-Mar
1 999-2002
June 2003-
June 2005
August 2005-
June 2006
September-
October 1999;
April-June
2000
May-
September
2005
03
Averaging
Time
1-h max
8-h moving
avg
24-h avg
8-h max
24-h avg
1-h max
Mean/Median
Concentration
(PPb)
28.2
31.6
17.5
62.9
26.3
74.5
Upper Percentile
Concentrations
(PPb)
75th: 36.0,
Max 70.0
Max: 86.3
90th: 26.8,
Max: 34.7
90th: 83.9,
Max: 105.9
75th: 35.3;
Max: 62.8
75th: 92.5;
Max: 165
*Note: Studies presented in order of first appearance in the text of this section.
ICAS = Inner City Asthma Study, NR = Not Reported, NCICAS = National Cooperative Inner-City Asthma Study.
Measurements at two sites established by investigators and located within 5 km of most subjects' residences.
""Concentrations converted from ug/m3 to ppb using the conversion factor of 0.51 assuming standard temperature (25°C) and
pressure (1 atm).
°Measured where subjects spent daytime hours.
In a majority of studies, including a large U.S. multicity study and several smaller
studies conducted in the United States, Mexico City, Mexico, and Europe, an
increase in ambient O3 concentration (various averaging times and lags) was
associated with a decrement in FEVi (Figure 6-7 [and Table 6-8]) or PEF (Figure 6-8
[and Table 6-9]) in children with asthma. Results were more variable for FEVi than
for PEF. In most studies, FEVi was measured by technicians whereas PEF was
measured by study subjects or their parents. However, associations with O3 also were
found with PEF measured by trained technicians (Romieu et al., 2004b; Thurston et
al., 1997), which are subject to less measurement error. Further, in some studies,
associations with FEVi were limited to specific subgroups. Some studies found that
increases in ambient O3 concentration were associated with greater lung function
variability, i.e., a deviation from a baseline level. These results pointed to
associations of O3 with poorer lung function, as indicated by a decrease from the
individual's mean lung function over the study period (Jalaludin et al., 2000), a
decrease in lung function over the course of the day (Lewis et al., 2005), or a
decrease in the lowest daily measurement (Just et al., 2002). Within many studies,
increases in O3 concentration were associated with decreases in lung function and
increases in respiratory symptoms at the same or similar lag (Just et al., 2002;
Mortimer et al.. 2002: GielenetaL 1997: Romieu etal.. 1997: Thurston et al.. 1997:
Romieu et al., 1996) (see Figure 6-12 [and Table 6-20]) for symptom results).
6-44
-------
Study
Liu etal. (2009)
Lewisetal. (2005)
Hoppeetal. (2003)
O3 Lag Subgroup
0
1
1 CSuser
With URI
1 Withoutasthma
With asthma
Barraza-Villarrealet al. 0-4 avg Withoutasthma
(2008)
Romieu etal. (2002) 1
Romieu etal. (2006)
With asthma
Placebo
Antioxidant
Placebo, moderate/severe asthma
Antioxidant, moderate/severe asthma
GSTP1 He/He orlle/Val
GSTP1 Val/Val
-10 -8 -6-4-20 2 4
Percent change in FEV., per unit increase in O3 (95% Cl)
Note: Results generally are presented in order of increasing mean ambient O3 concentration. CS = Corticosteroid, URI = Upper
respiratory infection. Effect estimates are from single-pollutant models and are standardized to a 40-ppb increase for 30-min or
1 -h max O3 concentrations, a 30-ppb increase for 8-h max or 8-h avg O3 concentrations, and a 20-ppb increase for 24-h avg O3
concentrations.
Figure 6-7 Percent change in FEVi in association with ambient O$
concentrations among children with asthma.
6-45
-------
Table 6-8
Study*
Liu et al.
(2009a)
Lewis et al.
(2005)
Hoppe et al.
(2003)
Barraza-
Villarreal et
al. (2008)
Romieu et
al. (2002)
Romieu et
al. (2006)
Percent change in FEVi in association with ambient O3
concentrations among children with asthma for studies
in Figure 6-7 plus others.
03
Location/ Averaging
Population Time
Windsor, ON, Canada
1 82 children with 24-h avg
asthma,
ages 9-1 4 yr
Detroit, Ml
86 children with
asthma, 8-h max
mean (SD)
age 9.1 (1.4)yr
Munich, Germany
43 people with
asthma,
ages 1 2-23 yr 30-mm max
(1-4 p.m.)
44 children without
asthma,
ages 6-8 yr
Mexico City, Mexico
208 children, 8-h avg
ages 6-1 4 yr
Mexico City, Mexico
1 58 children with 1 _n mgx
asthma,
ages 6-1 7 yr
Mexico City, Mexico
151 children with 1-hmax
asthma,
03
Lag
0
1
1
2
1
0-4
avg
1
1
Parameter Subgroup
CS user
Lowest daily with URI
FEVl CS user
With URI
Afternoon Without asthma
FEV, with asthma
Afternoon Without asthma
FVC with asthma
50 without asthma
FEV!
158 with asthma
Placebo
Antioxidant supplement
FEVi Placebo,
moderate/severe asthma
Antioxidant supplement,
moderate/severe asthma
GSTP1 lie/lie or Ile/Val
FEVl GSTP1 Val/Val
presented
Standardized
Percent Change
(95% Cl)a
-0.89 (-3.5, 1.8)
-0.44 (-2.4, 1 .6)
-1.9 (-10.4, 7.5)
-5.5 (-9.5, -1.5)
-7.3 (-12.3, -1.9)
-4.9 (-10.0, 0.48)
0.93 (-0.80, 2.7)
-0.56 (-4.6, 3.7)
-0.09 (-1.7, 1.6)
-3.5 (-5.9, -1.0)
-1.5 (-4.7, 1.7)
-0.1 2 (-2.0, 1.8)
-0.21 (-0.77, 0.36)
0.05 (-0.60, 0.69)
-1.1 (-2.0, -0.19)
-0.04 (-0.92, 0.83)
-0.51 (-1.1, 0.05)
0.50 (-0.25, 1.3)
mean age 9 yr
6-46
-------
Study*
Studies not
Dales et al.
(2009)
Rabinovitch
et al. (2004)
O'Connor et
al. (2008)
03
Location/ Averaging
Population Time
included in Figure 6-7b
Windsor, ON, Canada
1 82 children with ^ _n mgx
asthma,
ages 9-1 4 yr
Denver, CO
86 children with -|_n mgx
asthma,
ages 6-1 2 yr
Boston, MA;
Bronx, Manhattan NY;
Chicago, IL; Dallas,
TX, Seattle, WA;
Tucson, AZ 24'n av9
861 children with
asthma, mean (SD)
age 7.7 (2.0) yr
03
Lag Parameter
Evening
Q percent
predicted
FEV,
0-2 Morning
avg FEV, (ml)
. 5 Percent
predicted
9 FEV,
Standardized
Percent Change
Subgroup (95% Cl)a
-0.47 (-1.9, 0.95)
55 (-2.4, 108)
-0.41 (-1.0, 0.21)
'Includes studies in Figure 6-7. plus others
CS = corticosteroid, URI = Upper respiratory infection.
"Effect estimates are standardized to a 40-ppb increase for 30-min or 1-h max O3, a 30-ppb increase for 8-h max or 8-h avg O3, and
a 20-ppb increase for 24-h avg O3.
bResults not presented in Figure 6-7 because a different form of FEVi with a different scale was examined or because sufficient
data were not available to calculate percent change in FEN/!.
6-47
-------
Study Parameter
Gielenetal. (1997) Evening PEF
Morning PEF
Mortimeretal. (2002) Morning PEF
Mortimeretal. (2000) Morning PEF
Thurstonetal. (1997) PEF
Romieuetal. (2004b) FEF25.75o/0
Romieuetal. (1996) Evening PEF
Romieuetal. (1997) Evening PEF
O3Lag Subgroup
0
2
2
1
3
1-5avg All subjects
Normal BW
LowBW
No medication
CSuser
Placebo, GSTM1 null
Placebo, GSTM1 positive
Antioxidant, GSTM1 null
Antioxidant, GSTM1 positive
-10
-8
-2
Percent change in PEF or FEF25.75%
per unit increase in O3 (95% Cl)
Note: Results generally are presented in order of increasing mean ambient O3 concentration. BW = birth weight,
CS = Corticosteroid. Effect estimates are from single pollutant models and are standardized to a 40-ppb increase for 1-h max O3
concentrations and a 30-ppb increase for 8-h max or 8-h avg O3 concentrations.
Figure 6-8 Percent change in PEF or FEF2s-75% in association with ambient O$
concentrations among children with asthma.
6-48
-------
Table 6-9
Study*
Gielen et al.
(1997)
Mortimer et al.
(2002)
Mortimer et al.
(2000)
Thurston et al.
(1997)
Romieu et al.
(2004b)
Romieu et al.
(1996)
Romieu et al.
(1 997)
Percent change in PEF or FEF25-75% in association with ambient O$
concentrations among children with asthma for studies presented
in Figure 6-8 plus others.
Location/Population
Amsterdam, Netherlands
61 children with asthma,
ages 7-1 Syr
Bronx, East Harlem, NY;
Baltimore, MD; Washington,
DC; Detroit, Ml,
Cleveland, OH; Chicago, IL;
St. Louis, MO
846 children with asthma,
ages 4-9 yr
Bronx, East Harlem, NY;
Baltimore, MD; Washington,
DC; Detroit, Ml,
Cleveland, OH; Chicago, IL;
St. Louis, MO
846 children with asthma,
ages 4-9 yr
CT River Valley, CT
166 children with asthma,
ages 7-13 yr
Mexico City, Mexico
158 children with asthma,
mean age 9 yr
Northern Mexico City, Mexico
71 children with asthma,
ages 5-7 yr
Southern Mexico City, Mexic
0
65 children with asthma,
ages 5-13 yr
03
Averaging
Time
8-h max
8-h avg
(10 a.m.-
6 p.m.)
8-h avg
(10 a.m.-
6 p.m.)
1-h max
1 -h max
1 -h max
1-h max
03
Lag
0
2
2
1
3
1-5
avg
1-5
avg
0
1
0
2
0
2
Parameter Subgroup
Evening PEF
Evening PEF
Morning PEF
Morning PEF All subjects
Normal BW
LowBW(<5.5
Morning PEF Ibs.)
No medication
CS user
Intraday
change PEF
Placebo, GSTM1
null
Placebo, GSTM1
positive
I LI 2b-/i>%
Antioxidant,
GSTM1 null
Antioxidant, GST
M1 positive
Evening PEF
Evening PEF
Standardized
Percent
Change
(95% Cl)a
1.3 (-0.25, 2.9)
-1.3 (-2.8, 0.16)
-1.3 (-2.6, -0.08)
-0.1 2 (-0.76, 0.52)
-0.64 (-1.2, -0.10)
-1.2 (-2.1, -0.26)
-0.60 (-1 .6, 0.39)
-3.6 (-5.2, -2.0)
-1.1 (-3.0,0.84)
-1.2 (-2.5, 0.11)
-2.8 (-4.9, -0.59)
-2.3 (-4.2, -0.44)
-0.48 (-1 .7, 0.74)
-0.1 6 (-1.8, 1.6)
0.24 (-1.3, 1.8)
-0.1 7 (-0.79, 0.46)
-0.55 (-1.3, 0.19)
-0.52 (-1.0, -0.01)
-0.06 (-0.70, 0.58)
6-49
-------
Study*
Studies not
Jalaludin et al.
(2000)
Wiwatanadate
and
Trakultivakorn
(2010)
O'Connor et al
(2008)
Just et al.
(2002)
Location/Population
included in Figure 6-8b
Sydney, Australia
45 children with asthma and
mean (SD)
age9.6(1)yr
Chiang Mai, Thailand
31 children with asthma,
ages 4-1 1 yr
Boston, MA;
Bronx, Manhattan NY;
Chicago, IL; Dallas, TX,
Seattle, WA; Tucson, AZ
861 Children with asthma,
mean (SD)
age 7.7 (2.0) yr
Paris, France
82 children with asthma,
mean (SD)
age 10.9 (2.5) yr
03
Averaging Os
Time Lag Parameter
Daily
24-h avg deviation
1-hmax from mean
PEF
24-h avq ° Daily avg
** n avg g pEp (L/min)
Change in
~. . 1-5 percent
" avg predicted
PEF
„ 2 Percent
8-h avg " variability
avg pEp
Standardized
Percent
Change
Subgroup (95% Cl)a
-5.2 (-8.3, -2.2)°
-1.1 (-2.4,0.18)°
1.0 (-1.6, 3.6)
-2.6 (-5.2, 0)
-0.22 (-0.86, 0.43)
15.3(0,30.6)
'Includes studies InFigure 6-8. plus others
BW = birth weight, CS = corticosteroid, AHR = Airway hyperresponsiveness.
aEffect estimates are standardized to a 40-ppb increase for 1 -h max O3, a 30-ppb increase for 8-h max or avg O3, and a 20-ppb
increase for 24-h avg O3.
""Results are not presented in Figure 6-8 because a different form of PEF with a different scale was examined or because sufficient
data were not available to calculate percent change in PEF.
°Outcome defined as the normalized percent deviation from individual mean PEF during the study period. Quantitative results from
generalized estimating equations were provided only for models that included PM10 and NO2.
The most geographically representative data were provided by the large, multi-U.S.
city National Cooperative Inner City Asthma Study (NCICAS) (Mortimer et al.,
2002: Mortimer et al.. 2000) and Inner-City Asthma Study (ICAS) (O'Connor et al..
2008). Although the two studies differed in the cities, seasons, racial distribution of
subjects, and lung function indices examined, results were fairly similar. In ICAS,
which included children with asthma and atopy (i.e., allergic sensitization) and year-
round examinations of lung function, a 20-ppb increase in the lag 1-5 average of
24-h avg O3 was associated with a 0.41-point decrease in percent predicted FEVi
(95% CI: -1.0, 0.21) and a 0.22-point decrease in percent predicted PEF (95% CI:
-0.86, 0.43) (O'Connor et al.. 2008).
Increases in lag 1-5 avg O3 (8-h avg, 10 a.m.-6 p.m.) also were associated with
declines in PEF in NCICAS, which included different U.S. cities, summer-only
measurements, larger proportions of Black and Hispanic children, and fewer subjects
with atopy (79%) (Mortimer et al., 2002). Ozone concentrations lagged 3 to 5 days
were associated with larger PEF decrements than were O3 concentrations lagged 1 to
2 days (Figure 6-8 [and Table 6-91). NCICAS additionally identified groups
potentially at increased risk of O3-associated PEF decrements, namely, males,
6-50
-------
children of Hispanic ethnicity, children living in crowded housing, and as indicated
in Figure 6-8 (and Table 6-9), children with birth weight <5.5 Ibs (Mortimer et al.,
2000). Somewhat paradoxically, O3 was associated with a larger decrease in PEF
among subjects taking cromolyn, medication typically used to treat asthma due to
allergy, but a smaller decrease among subjects with allergic sensitization (as
determined by skin prick test). NCICAS also indicated robust associations with
consideration of other sources of heterogeneity. Except for Baltimore, MD, effect
estimates were similar across the study cities (1.1 to 1.7% decrease in PEF per
30-ppb increase in lag 1-5 avg of 8-h avg O3). Results were similar with O3 averaged
from all available city monitors and concentrations averaged from the three monitors
closest to subjects' ZIP code centroid (1.2% and 1.0% decrease in PEF, respectively,
per 30-ppb increase in O3). At concentrations <80 ppb, a 30-ppb increase in lag 1-5
of 8-h avg O3 was associated with a 1.4% decrease (95% CI: -2.6, -0.21) in PEF,
(Mortimer et al.. 2002) which was similar to the effect estimated for the full range of
O3 concentrations (Figure 6-8 [and Table 6-91). In a study of children with asthma in
the Netherlands, Gielen et al. (1997) estimated similar effects on PEF for the full
range of 8-h max O3 concentrations and concentrations <51 ppb.
Several but not all controlled human exposure studies have reported slightly larger
O3-induced FEVi decrements in adults with asthma than adults without asthma
(Section 6.2.1.1). However, in the few epidemiologic studies that compared children
with and without asthma, evidence did not conclusively indicate that children with
asthma were at increased risk of O3-associated lung function decrements. Hoppe et
al. (2003) and Jalaludin et al. (2000) generally found larger O3-associated
decrements in FVC and PEF, respectively, in children with asthma; whereas
Raizenne et al. (1989) did not consistently demonstrate differences between campers
with and without asthma. In their study of children in Mexico City, Mexico, Barraza-
Villarreal et al. (2008) estimated larger O3-associated decreases in children without
asthma; however, 72% of these children had atopy. These findings indicate that
children with atopy, who also have airway inflammation and similar respiratory
symptoms, may experience respiratory effects from short-term ambient O3 exposure.
As shown in Figure 6-7 (and Table 6-81) and Figure 6-8 (and Table 6-9), lung
function decrements in children with asthma mostly ranged from <1% to 2% per unit
increase in ambient O3 concentration1. Larger magnitudes of decrease, were found in
children with asthma who were using CS, had a concurrent upper respiratory
infection (UPJ), were GSTM1 null, had airway hyperresponsiveness, or had
increased outdoor exposure (Romieu et al., 2006; Lewis et al., 2005; Romieu et al.,
2004b; Jalaludin et al., 2000) than among children with asthma overall (Barraza-
Villarreal et al.. 2008: Lewis et al.. 2005: Delfino et al.. 2004: Romieu et al.. 2002).
For example, Jalaludin et al. (2000) estimated a -5.2% deviation from mean FEVi
per 20-ppb increase in 24-h avg O3 concentration among children with asthma and
airway hyperresponsiveness and a much smaller -0.71% deviation among children
with asthma without airway hyperresponsiveness. In a group of 86 children with
asthma in Detroit, MI, Lewis et al. (2005) reported that associations between ambient
1 Effect estimates were standardized to a 40-, 30-, and 20-ppb increase for 1-h max, 8-h max or 8-h avg, and 24-h avg O3.
6-51
-------
O3 concentration and FEVi were confined largely to children with asthma who used
CS or had a concurrent URI, 7.3% and 4.9% decreases, respectively, in the mean of
lowest daily FEVi per 30-ppb increase in 8-h max ambient O3 concentration.
Heterogeneity in response to ambient O3 exposure also was demonstrated by
observations that some children with asthma experienced larger O3-associated lung
function decrements than the population mean effect estimate. Similar observations
were made in controlled human exposure studies (Section 6.2.1.1). Mortimer et al.
(2002) found that for a 30-ppb increase in lag 1-5 avg of 8-h avg O3, there was a
30% ([95% CI: 4%, 61%]) higher incidence of >10% decline in PEF. Likewise,
Hoppe et al. (2003) found that while the percentages of O3-associated lung function
decrements were variable and small, 47% of children with asthma experienced a
>10% decline in FEVi, FVC, or PEF or 20% increase in airway resistance on days
with 30-min (1-4 p.m.) max ambient O3 concentrations >50 ppb relative to days with
<40ppbO3.
Effect Modification
Effect modification by corticosteroid use
In controlled human exposure studies, CS treatment of subjects with asthma
generally has not prevented O3-induced FEVi decrements (Section 6.2.1.1).
Epidemiologic evidence is equivocal, with findings that use of inhaled CS attenuated
(Hernandez-Cadena et al., 2009), increased (Lewis et al., 2005), and did not affect
(Mortimer et al., 2000) ambient O3-associated lung function decrements. In winter-
only studies, consideration of CS use largely did not influence associations between
ambient O3 and various lung function indices (Liu et al., 2009a; Rabinovitch et al.,
2004). Similarly equivocal epidemiologic evidence was found for modification of
associations with respiratory symptoms (Section 6.2.4.1). The assessment of effect
modification by CS use has been hampered by differences in the severity of asthma
among CS users and the definition of CS use. Additionally, investigators did not
assess adherence to reported CS regimen, and misclassification of CS use may bias
findings. For example, Mortimer et al. (2000) classified children by no or any CS use
at baseline but did not measure daily use during the study period. Lewis et al. (2005)
defined CS use as use for at least 50% of study days and estimated larger
O3-associated FEVi decrements among CS users (Figure 6-7 [and Table 6-81) than
among CS nonusers (quantitative results not reported). In this study, most children
with moderate to severe asthma (91%) were classified as CS users. However, CS
users had a higher percent predicted FEVi. In contrast, Hernandez-Cadena et al.
(2009) observed larger O3-related decrements in FEVi among the 60 CS nonusers
than among the 25 CS users. A definition for CS use was not provided; however,
children with persistent asthma were included among the group of CS nonusers.
Thus, across studies, both CS use and nonuse have been used to indicate more
severe, uncontrolled asthma.
6-52
-------
Effect modification by antioxidant capacity
Ozone is a powerful oxidant whose secondary oxidation products have been
described to initiate the key modes of action that mediate decreases in lung function,
including the activation of neural reflexes (Section 5.3.2). Additionally, O3 exposure
of humans and animals has induced changes in the levels of antioxidants in the ELF
(Section 5.3.3). These observations provide biological plausibility for diminished
antioxidant capacity to increase the risk of O3-associated respiratory effects and for
augmented antioxidant capacity to decrease risk.
Antioxidant supplementation
Controlled human exposure studies have demonstrated the protective effects of a-
tocopherol (vitamin E) and ascorbate (vitamin C) supplementation on O3-induced
lung function decrements (Section 6.2.1.1). and an epidemiologic study of children
with asthma conducted in Mexico City, Mexico, produced similar findings.
Particularly among children with moderate to severe asthma, an increase in ambient
O3 concentration was associated with a smaller decrease in FEVi in the group
supplemented with vitamin C and E as compared with the placebo group (Romieu et
al., 2002) (Figure 6-7 [and Table 6-81). Romieu et al. (2009) also demonstrated an
interaction between dietary antioxidant intake and ambient O3 concentrations by
finding that the main effect of diet was modified by ambient O3 concentrations. Diets
high in antioxidant vitamins and/or omega-3 fatty acids protected against FEVi
decrements at 8-h max O3 concentrations > 38 ppb. Results for the main effect of O3
on FEVi or effect modification by diet were not presented.
Genetic polymorphisms
Antioxidant capacity also can be characterized by variants in genes encoding oxidant
metabolizing enzymes with altered enzymatic activity. A potential role for such
genetic variants in modifying O3-associated health effects is biologically plausible
given the well-characterized evidence for the secondary oxidation products of O3
mediating downstream effects and has been indicated in some epidemiologic studies.
Specifically, ambient O3-associated FEF25-75% decrements were larger among
children with asthma with the GSTM1 null genotype, which is associated with lack
of oxidant metabolizing activity (Romieu et al.. 2004b). The difference in association
between GSTM1 null and positive subjects was minimal in children supplemented
with antioxidant vitamins (Figure 6-8 [and Table 6-91). Controlled human exposure
studies have not consistently found larger O3-induced lung function decrements in
GSTM1 null subjects (Section 6.2.1.1). Effect modification by GSTP1 variants is
less clear. Romieu et al. (2006) observed larger O3-associated decreases in FEVi in
children with asthma with the GSTP1 lie/lie or Ile/Val variant, which are associated
with relatively higher oxidative metabolism activity (Figure 6-7 [and Table 6-8]).
An increase in ambient O3 concentration was associated with an increase in FEVi
among children with the GSTP1 Val/Val variant, which is associated with reduced
oxi dative metabolism. Rather than reflecting effect modification by the GSTP1
variant, these results may reflect effect modification by asthma severity, as 77% of
6-53
-------
subjects with the GSTP1 lie/lie genotype had moderate to severe asthma. In support
of this alternate hypothesis, another analysis of the same cohort indicated a larger
Os-associated decrement in FEVi among children with moderate to severe asthma
than among all children with asthma (Romieu et al.. 2002).
Exposure Measurement Error
Across the studies of children with asthma, lung function decrements were associated
with ambient O3 concentrations assigned to subjects using various exposure
assessment methods. As described in Section 4.3.3. exposure measurement error due
to use of ambient concentrations measured at central sites has varied, depending on
the population and season examined. Because there are a limited number of studies
of each method, it is difficult to conclude that a particular method of exposure
assessment produced stronger results.
Seasonal differences have been observed in the personal-ambient O3 relationship
(Section 4.3.3); however, in children with asthma, O3-associated lung function
decrements were found in studies conducted in summer months and over multiple
seasons. Lung function was associated with O3 measured on site of subjects' daytime
hours in summer months (Hoppe et al., 2003; Thurston et al., 1997), factors that have
contributed to higher personal-ambient O3 ratios and correlations. Many year-round
studies in Mexico City, Mexico (Romieu et al.. 2006; 2004b; 2002; 1997; 1996). and
a study in Detroit, MI (Lewis et al.. 2005) found associations with O3 measured at
sites within 5 km of children's home or school. Children with asthma examined by
Romieu et al. (2006); (2004b; 2002) had a personal-ambient ratio and correlation for
48- to 72-h avg O3 concentrations of 0.17 and 0.35, respectively (Ramirez-Aguilar et
al.. 2008). These findings indicate that the effects of personal O3 exposure on lung
function decrements may have been underestimated in the children in Mexico City.
Associations were found with O3 concentrations averaged across multiple
community monitoring sites (O'Connor et al.. 2008; Just et al.. 2002; Mortimer et al..
2002; Jalaludin et al.. 2000) and measured at a single site (Gielen et al.. 1997). which
may be attributable to observations of high temporal correlation among. O3
concentrations measured at multiple sites within a region (Darrow et al.. 201 la; Gent
et al.. 2003).
Studies of children with asthma restricted to winter months provided little evidence
of an association between various single- and multi-day lags of ambient O3
concentration and lung function decrements with several observations of
O3-associated increases in lung function (Dales et al.. 2009; Liu et al.. 2009a;
Rabinovitch et al.. 2004). One explanation for these results may be lower indoor than
outdoor O3 concentrations, variable indoor to outdoor ratios, and lower correlations
between personal and ambient O3 concentrations in non-summer months (Sections
4.3.2 and 4.3.3). As noted for other respiratory endpoints such as respiratory hospital
admissions, ED visits, and mortality, associations with O3 generally are lower in
colder seasons.
6-54
-------
Adults with Respiratory Disease
Relative to studies in children with asthma, studies of adults with asthma or COPD
have been limited in number. Details from these studies regarding location, time
period, and ambient O3 concentrations are presented in Table 6-10. Increases in
ambient O3 concentration were not consistently associated with lung function
decrements in adults with respiratory disease. Several different exposure assessment
methods were used, including monitoring personal exposures (Delfino et al., 1997),
monitoring on site of outdoor activity (Girardot et al., 2006; Korrick et al., 1998),
and using measurements from one (Peacock et al., 2011; Wiwatanadate and
Liwsrisakun, 2011; Thaller et al., 2008; Ross et al., 2002) to several central monitors
(Khatri et al., 2009; Lagorio et al., 2006; Park et al., 2005a). There was not a clear
indication that differences in exposure assessment methodology contributed to
inconsistencies in findings.
6-55
-------
Table 6-10 Mean and upper percentile concentrations of O3 in epidemiologic
studies of lung function in adults with respiratory disease.
Study*
Delfino et al.
(1 997)
Girardot et al.
(2006)
Korrick et al.
(1 998)
Peacock et al.
(2011)
Wiwatanadate and
Liwsrisakun
(2011)
Thaller etal.
(2008): Brooks
(2010)
Ross et al. (2002)
Khatri et al. (2009)
Lagorio et al.
(2006)
Location
Alpine, CA
Great Smoky
Mountain NP,
TN
Mt.
Washington,
NH
London,
England
Chiang Mai,
Thailand
Galveston, TX
East Moline, IL
Atlanta, GA
Rome, Italy
Study Period
May- July 1994
August-October
2002 June-August
2003
Summer 1991, 1992
All-year 1995-1997
August 2005-June
2006
Summer 2002-2004
April-October 1994
May-September
2003, 2005, 2006
May-June,
November-
December 1999
Os Averaging
Time
12-h avg
personal
(8 a.m. -8 p.m.)
Hike-time avg
(2-9 h)
Hike-time avg
(2-1 2 h)
8-h max
24-h avg
1 -h max
8-h avg
8-h max
24-h avg
Mean/Median
Concentration (ppb)
18
48. 1a
40a
15.5
17.5
35 (median)
41.5
With asthma: 61
(median)b
No asthma: 56
(median)b
Spring: 36.2°
Winter: 8.2°
Upper Percentile
Concentrations (ppb)
90th: 38
Max: 80
Max: 74.2a
Max: 74a
Autumn/Winter Max: 32
Spring/Summer Max: 74
90th: 26.8
Max: 34.7
Max: 118
Max: 78.3
75th (with asthma): 74b
75th (no asthma): 64b
Overall max: 48.6°
Park etal. (2005a) Incheon, Korea March-June 2002
24-h avg
Dust event days: 23.6
Control days: 25.1
NR
*Note: Studies presented in order of first appearance in the text of this section.
NR = Not reported.
alndividual-level estimates calculated from concentrations measured in different segments of hiking trail.
blndividual-level estimates calculated based on time spent in the vicinity of various O3 monitors.
""Concentrations converted from ug/m3 to ppb using the conversion factor of 0.51 assuming standard temperature (25°C) and
pressure (1 atm).
Comparisons of adults with asthma (8-18% of study population) and without asthma
did not conclusively demonstrate that adults with asthma had larger ambient
O3-associated lung function decrements. Several studies examined on-site or central-
site ambient O3 concentrations measured while subjects were outdoors. Ambient O3
measured during time spent outdoors has been closer in magnitude and more
correlated with personal exposures (Section 4.3.3). In a panel study of lifeguards
(ages 16-27 years) in Galveston, TX, a larger O3-associated decrement in FEVi/FVC
was found among the 16 lifeguards with asthma (-1.6% [95% CI: -2.8, -0.4] per
40 ppb increase in 1-h max O3) than among the 126 lifeguards without asthma
(-0.40% [95% CI: -0.80, 0] per 40-ppb increase in 1-h max O3) (Brooks. 2010).
In Korrick et al. (1998), hikers with a history of asthma or wheeze had larger
6-56
-------
O3-associated lung function decrements (e.g., -4.4% [95% CI: -7.5, -1.2] in FEVi per
30-ppb increase in 2-12 h avg O3). In contrast, Girardot et al. (2006) generally did
not find O3-associated lung function decrements in hikers with or without respiratory
disease history. In a cross-sectional study of 38 adults with asthma and 13 adults
without asthma, Khatri et al. (2009) used central site O3 measurements but aimed to
account for spatial variability by calculating an average of concentrations measured
at sites closest to each subject's location during each hour. Investigators reported a
larger O3-associated decrease in percent predicted FEVi/FVC in the 38 subjects with
atopy (with or without asthma) (-12 points [95% CI: -21, -3] per 30-ppb increase in
lag 2 of 8-h max O3) than in subjects with asthma (-4.7 points [95% CI: -12, 2.3]).
Among adults with asthma, O3 was associated with an increase in FEVi.
In panel studies that exclusively examined adults with asthma, increases in ambient
O3 concentrations, across the multiple lags examined, generally were associated with
increases in lung function (Wiwatanadate and Liwsrisakun, 2011; Lagorio et al.,
2006; Park et al., 2005a). These studies were conducted in Europe and Asia during
periods of low ambient O3 concentrations, including one conducted in Korea during
a period of dust storms (Park et al., 2005a).
Some studies included children and adults with asthma. Among subjects ages 9-
46 years (41% adults) in Alpine, CA with low personal 12-h avg O3 exposures (55%
samples below limit of detection) and a majority of sampling hours spent indoors
(mean 71%), Delfino et al. (1997) reported that neither increases in 12-h avg
personal exposure nor increases in ambient O3 concentration were associated with
decreases in PEF. Ross et al. (2002) examined subjects ages 5-49 years (proportion
of adults not reported) in East Moline, IL and found that a 20-ppb increase in lag 0
(of 24-h avg O3) was associated with a 2.6 L/min decrease (95% CI: -4.3, -0.90) in
evening PEF. In this population with asthma, an increase in lag 0 ozone also was
associated with an increase in symptom score.
Controlled human exposure studies have found diminished, statistically
nonsignificant O3-induced lung function responses in older adults with COPD
(Section 6.2.1.1). Similarly, epidemiologic studies do not provide strong evidence
that short-term increases in ambient O3 exposure result in lung function decrements
in adults with COPD. Inconsistent associations were reported for PEF, FEVi, and
FVC in a study that followed 94 adults with COPD (ages 40-83 years) in London,
England daily over two years (Peacock et al.. 2011). For example, an increase in lag
1 of 8-h max O3 was associated with a decrease in PEF in an analysis of summer
1996 (-1.7 L/min [95% CI: -3.1, -0.39] per 30-ppb increase O3), but the association
was near null and imprecise in summer 1997 (-0.21 L/min [95% CI: -2.4, 2.0]).
Further, in this study, an increase in ambient O3 concentration was associated with
lower odds of a large PEF decrement (OR for a >20% drop from an individual's
median value: 0.89 [95% CI: 0.72, 1.10] per 30-ppb increase in lag 1 of 8-h max O3)
and was not consistently associated with increases in respiratory symptoms (Peacock
et al.. 2011). Inconsistent associations also were reported in a small panel study of 11
adults with COPD (mean age 67 years) in Rome, Italy (Lagorio et al.. 2006).
6-57
-------
Populations Not Restricted to Individuals with Asthma
Several studies have examined associations between ambient O3 concentrations and
lung function in groups that included children with and without asthma; however, a
limited number of studies have examined groups of children or adults restricted to
healthy individuals. Details from studies not restricted to individuals with asthma
regarding location, time period, and ambient O3 concentrations are presented in
Table 6-11.
6-58
-------
Table 6-11 Mean and upper percentile concentrations of O3 in epidemiologic
studies of lung function in populations not restricted to individuals
with asthma.
Study*
Avoletal. (1998b)
Hoppe et al. (2003)
Chenetal. (1999)
Goldetal. (1999)
Ward et al. (2002)
Ulmeretal. (1997)
Linnetal. (1996)
Scarlett et al.
(1996)
Neuberger et al.
(2004)
Alexeeffetal.
(2008): (2007)
Steinvil et al.
(2009)
Naeher et al.
(1999)
Sonetal. (2010)
Location
6 southern CA
communities
Munich,
Germany
3 Taiwan
communities
Mexico City,
Mexico
Birmingham
and Sandwell,
England
Freudenstadt
and Villingen,
Germany
Rubidoux,
Upland,
Torrance, CA
Surrey, England
Vienna, Austria
Greater Boston,
MA; MAS
Tel Aviv, Israel
Multiple
communities,
VA
Ulsan, Korea
Study Period
Spring and
summer 1994
Summers
1 992-1 995
May 1 995-
January 1996
January-
November 1991
January-March,
May-July 1997
March-October
1994
September-June
1 992-1 994
June-July 1994
June-October
1999, January-
April 2000
January 1995-
June 2005
September 2002-
November2007
May-September
1 995-1 996
All-year, 2003-
2007
O3 Averaging
Time
24-h avg personal
30-min max (1-
4 p.m.)
1-h max (8a.m.-
6 p.m.)
24-h avg
24-h avg
30-min avg
24-h avg personal
24-h avg central
site
8-h max
NR
48-h avg
8-h avg
(10a.m.-6p.m.)
24-h avg
8-h max
Mean/Median
Concentration
(PPb)
NR
High 03 days: 70.4a
Control O3 days: 29.8a
NR
52.0a
Winter median: 13.0
Summer median: 22.0
Freudenstadt median:
50.6
Villingen median: 32.1
5
23
50.7a
NR
24.4b
41.1
34.9
35.9
(avg of 13 monitors)
Upper Percentile
Concentrations
(PPb)
NR
Max (high O3 days):
99a
Max (control O3 days):
39a
Max: 110.3a
Max: 103a
Winter Max: 33
Summer Max: 41
Freudenstadt 95th:
89.8
Villingen 95th: 70.1
Max: 16
Max: 53
Max: 128a
NR
NR
75th: 48.7
Max: 72.8
Max: 56.6
Max: 59.5
*Note: Studies presented in order of first appearance in the text of this section.
NR = Not Reported, NAS = Normative Aging Study.
aMeasured at subjects' schools where lung function was measured.
""Measured at central monitoring sites established by investigators. Concentrations were averaged across four monitors.
Children
Based on studies available at the time of the 2006 O3 AQCD, evidence consistently
links increases in ambient O3 concentration with decrements in FEVi and PEF in
children (U.S. EPA, 2006b) (Figure 6-9 [and Table 6-121). These associations were
6-59
-------
found with personal O3 exposures (Avol et al., 1998b), ambient O3 measured at
children's schools where lung function was measured (Hoppe et al., 2003; Chen et
al., 1999; Gold et al., 1999), and ambient O3 measured at sites within the community
(Ward et al., 2002; Ulmeretal, 1997; Linn et al., 1996). Among children in
California who spent a mean 2-3 hours per day outdoors and whose personal-ambient
O3 correlation was 0.28 across multiple seasons, Avol et al. (1998b) found slightly
larger O3-associated decrements in FEVi and FVC for 24-h avg personal exposures
than for 1-h max ambient measurements (Figure 6-9 [and Table 6-121). The effect
estimates for personal exposures were similar in magnitude to those found in other
studies of children for ambient O3 measured at schools (Hoppe et al.. 2003; Chen et
al.. 1999). In another study of children in California, Linn et al. (1996) did not
present results for personal O3 exposures but found FEVi decrements in association
with increases in ambient O3 concentration in children who spent 1-2 hours per day
outdoors and whose personal-ambient correlation was 0.61. Because of between-
study heterogeneity in populations and ambient O3 concentrations examined, it is
difficult to assess how the method of exposure assessment may have influenced
findings.
Study
Parameter
O3Lag
Linn etal. (1996) Intraday change FEV., 0
Intraday change FVC
Hoppe etal. (2003) Afternoon FEV! 0
Afternoon FVC
Scarlettetal. (1996) FEV075
FVC
Chenetal. (1999)
1
FVC
Avol etal. (1998)a Intraday change FEV! 0 Personal
Ambient
Intraday change FVC Personal
Ambient
-10
-6
-2
0
Percentchangein FEV., or FVC per unit
increase in O3 (95% Cl)
Note: Results generally are presented in order of increasing mean ambient O3 concentration. Effect estimates are from single-
pollutant models and are standardized to a 40-, 30-, and 20-ppb increase for 1-h (or 30-min) max, 8-h max, and 24-h avg O3
concentrations, respectively.
"The 95% Cl was constructed using a standard error that was estimated from the p-value.
Figure 6-9 Percent change in FEVi or FVC in association with ambient Os
concentrations in studies of children in the general population.
6-60
-------
Table 6-12 Percent change in FEVi or FVC in association with ambient Os
concentrations in studies of children in the general population
presented in Figure 6-9 plus others.
Study*
Linn et al.
(1996)
Hoppe et
al. (2003)
Scarlett et
al. (1996)
Chen et al.
(1999)
Avol et al.
(1998b)
Location/ Population
3 southern CA communities
269 children, 4th and 5th grades
Munich, Germany
44 children, ages 6-8 yr
Surrey, England
154 children, ages 7-1 1 yr
3 Taiwan communities
941 children, ages 8-1 3 yr
3 southern CA communities
195 children, ages 10-1 2 yr
Studies of children not included in Figure 6-9
Ulmer et al.
(1 997)
Ward et al.
(2002)
Goldetal.
(1999)
Freudenstadt and Villingen,
Germany
1 35 children, ages 8-1 0 yr
Birmingham and Sandwell, England
162 children, age 9 yr
Mexico City, Mexico
40 children, ages 8-1 1 yr
O3 Averaging
Time
24-h avg
30-min max
(1 -4p.m.)
8-h max
1 -h max
(8 a.m. -6 p.m.)
24-h avg
personal
1-h max ambient
24-h avg
personal
1-h max ambient
c
30-min max
24-h avg
24-h avg
03
Lag
0
0
1
1
0
1
0
2
0
1-10
avg
Parameter
Intraday change FEN/!
Intraday change FVC
Afternoon FEN/!
Afternoon FVC
FEVo.75
FVC
FEV,
FVC
Intraday change FEVi
Intraday change FEV!
Intraday change FVC
Intraday change FVC
FEV, (ml)
Daily deviation from
mean morning PEF
(L/min)
Intraday change PEF
(% change)
Standardized
Percent Change
(95% Clf
-0.58 (-1.0, -0.13)
-0.21 (-0.62, 0.20)
-0.1 4 (-2.7, 2.5)
-1.4 (-3.9, 1.2)
-0.04 (-0.32, 0.23)
0.06 (-0.21, 0.33)
-1.5 (-2.8, -0.12)
-1 .6 (-2.9, -0.33)
-0.85 (-2.1,0.42)"
-0.49 (-1 .5, 0.57)b
-1.0 (-2.0, 0)b
-0.50 (-1 .3, 0.35)b
-57 (-102,13)"
-3.2 (-8.3, 2.0)d
-6.7 (-12, -1.4)d
-0.47 (-1.1, 0.11)
-3.8 (-6.7, -0.94)
Includes studies in Figure 6-9, plus others.
aEffect estimates are standardized to a 40-, 30-, and 20-ppb increase for 1-h (or 30-min) max, 8-h max, and 24-h avg O3,
respectively.
"The 95% Cl was constructed using a standard error that was estimated from the p-value.
°Results are not presented in Figure 6-9 because sufficient data were not available to calculate percent change in FEVi, or PEF was
analyzed.
dEffect estimates are from analyses restricted to summer months.
In the limited number of studies that examined only healthy children, increases in
ambient O3 concentration were associated with decreases (Hoppe et al.. 2003) or no
change in lung function (Neuberger et al.. 2004). Several studies that included small
proportions (4-10%) of children with history of respiratory disease or symptoms
found associations between increases in ambient O3 concentration and lung function
decrements (ChenetaL 1999: Ulmer et al.. 1997: Scarlett et al.. 1996). Based on
analysis of interaction terms for O3 concentration and asthma/wheeze history, Avol
et al. (1998b) and Ward et al. (2002) did not find differences in O3-associated lung
function decrements between children with history of asthma or wheeze and healthy
children. Combined, these lines of evidence suggest that the ambient O3-associated
6-61
-------
lung function decrements found in children overall were not solely due to effects in
children with asthma, and that increases in ambient O3 exposure may decrease lung
function in healthy children.
Among the studies of children, the magnitudes of decrease in lung function per unit
increase in ambient O3 concentration1 ranged from <1 to 4%, a range similar to that
estimated in children with asthma. Comparable data were not adequately available to
assess whether mean lung function differed between groups of children with asthma
and healthy children. In contrast with studies of children with asthma, studies of
children in the general population did not consistently find both O3-associated
decreases in lung function and O3-associated increases in respiratory symptoms. For
example, Gold et al. (1999) found O3-associated decreases in PEF and increases in
phlegm; however, the increase in phlegm was associated with lag 1 O3
concentrations whereas the PEF decrement was found with single-day lags 2 to 4 of
O3. Also, O3 was weakly associated with cough and shortness of breath among
children in England (Ward et al., 2002) and was associated with a decrease in
respiratory symptom score among children in California (Linn et al., 1996).
Adults
Compared with children, in a smaller body of studies, O3 was less consistently
associated with lung function decrements in populations of adults not restricted to
those with asthma (Table 6-13). In a study that included only healthy adults,
increases in ambient O3 concentration were associated with decreases and increases
in lung function across the various lags of exposure examined (Steinvil et al.. 2009).
Contrasting results also were found in studies of older adults (Alexeeff et al.. 2008:
Alexeeff et al.. 2007: Hoppe et al.. 2003).
1 Effect estimates were standardized to a 40-, 30-, and 20-ppb increase for 1-h max (or 30-min max), 8-h max, and 24-h avg O3.
6-62
-------
Table 6-13 Associations between ambient O3 concentration and lung function
in studies of adults.
Study3
Son et al.
(2010)
Steinvil et
al. (2009)
Naeher et
al. (1999)
Hoppe et
al. (2003)
Alexeeff et
al. (2008)
Alexeeff et
al. (2007)
Location/Population
Ulsan, Korea
2,102 children and adults,
ages 7-97 yr
Tel Aviv, Israel
2,380 healthy adults,
mean age 43 yr,
75th percentile: 52 yr
Multiple communities, VA
473 women,
ages 1 9 - 43 yr
Munich, Germany
41 older adults,
ages 69 - 95 yr
Greater Boston, MA; NAS
1,01 5 older adults,
mean (SD) age: 68.8
(7.2) yr at baseline
Greater Boston, MA; NAS
904 older adults,
mean (SD)age: 68.8
(7.3) yr at baseline
03
Averaging O3
Time Lag Parameter
Change in
8-h- °fg Sdfcted
8-h avg °
(10a.m. FEV, (ml)
- 6 p.m.)
avg
0
24-hav9 0-2 (L/min)
avg
rrr ? -— "
(1-4 p.m.) 1 pEVi
„. , 0-1 % change in
av9 avg FEV!
24 h ava °~1 % Cnan9e in
9 avg FEV!
O3 Assessment
Method/Subgroup
All monitor avg
Nearest monitor
IDW
Kriging
GSTP1 lie/lie
GSTP1 Ile/Val or
Val/Val
BMI <30
BMI > 30
No AHR
AHR
BMI > 30 and AHR
Standardized
Effect
Estimate
(95% Cl)b
-1 .4 (-2.7, -0.08)
-0.76 (-1.8,
0.25)
-1.1 (-2.2, 0.05)
-1.4 (-2.6, -0.11)
60(0,120)
141 (49,234)
-1 .7 (-3.4, 0.03)
-3.0 (-4.4, -1 .7)
0.75 (-2. 1,3.7)
1.2 (-1.3, 3.6)
-1 .0 (-2.2, 0.20)
-2.3 (-3.5, -1 .0)
-1.5 (-2.5, -0.51)
-3.5 (-5.1, -1.9)
-1 .7 (-2.7, -0.73)
-4.0 (-6.2, -1 .8)
-5.3 (-8.2, -2.3)
IDW= Inverse distance weighting, MAS = Normative Aging Study, BMI = Body mass index, AHR = airway hyperresponsiveness.
"Results generally are presented in order of increasing mean ambient O3 concentration.
bEffect estimates are standardized to a 40-ppb increase for 30-min max O3, 30-ppb increase for 8-h max or 8-h avg O3, and 20-ppb
increase for 24-h avg O3.
Despite mixed results overall, studies that found ambient O3-associated lung function
decrements in adults used various exposure assessment methods with potentially
varying degrees of measurement error. These methods included the average of
multiple intra-city monitors, nearest monitor, estimates from spatial interpolation
(Son et al., 2010), average of monitors in multiple towns (Alexeeff et al., 2008;
Alexeeff et al., 2007), and one site for multiple towns (Naeher et al., 1999). In a large
cross-sectional study, conducted in 2,102 children and adults (mean age: 45 years)
living near a petrochemical plant in Ulsan, Korea, Son et al. (2010) did not find a
consistent difference in the magnitude of association with lung function among
ambient O3 concentrations averaged across 13 city monitors, concentrations from the
nearest monitor, inverse distance-weighted concentrations, and estimates from
kriging across the various lags examined (Table 6-13). Ozone concentrations were
similar (<10% difference) and highly correlated (r = 0.84 - 0.96) among the methods.
6-63
-------
Although the health status of subjects was not reported, the study population mean
percent predicted FEVi was 82.85%, indicating a large proportion of subjects with
underlying airway obstruction. Results from this study were not adjusted for
meteorological factors and thus, confounding cannot be ruled out. Importantly, the
similarities among exposure assessment methods in Son et al. (2010) may apply
mostly to populations living within the same region of a city. The majority of women
examined by Naeher et al. (1999) lived >60 miles from the single available central
site monitor. However, in the nonurban (southwest Virginia) study area, O3
concentrations may be more spatially homogeneous (Section 4.6.2.1). and the
concentrations measured at the single site may capture temporal variability in
ambient exposures.
The inconsistent epidemiologic findings for older adults parallel observations from
controlled human exposure studies (Section 6.2.1.1). In a study that followed adults
ages 69-95 years during several summers in Germany, Hoppe et al. (2003) did not
find decreases in lung function in association with ambient O3 measured at subjects'
retirement home. However, recently, the Normative Aging Study found decrements
in FEVi and FVC in a group of older men (mean [SD] age = 68.9 [7.2] years at first
lung function measurement) in association with ambient O3 concentrations averaged
from four town-specific monitors (Alexeeff et al., 2008), which may less well
represent spatial heterogeneity in ambient O3 exposures. Among all subjects, who
were examined once every three years for ten years, associations were found with
several lags of 24-h avg O3 concentration, i.e., 1- to 7-day avg (Alexeeff et al.,
2008). Additionally, larger effects were estimated in adults with airway
hyperresponsiveness, higher BMI (> 30), and GSTP1 Ile/Val or Val/Val genetic
variants (Val/Val variant produces enzyme with reduced oxidative metabolism
activity) (Alexeeff et al.. 2008: Alexeeff et al.. 2007) (Table 6-13). Larger O3-related
decrements in FEVi and FVC also were observed in subjects with long GT
dinucleotide repeats in the promoter region of the gene for the antioxidant enzyme
heme oxygenase-1 (Alexeeff et al.. 2008). which has been associated with reduced
inducibility (Hiltermann et al.. 1998). In this cohort, O3 also was associated with
decreases in lung function in adults without airway hyperresponsiveness and those
with BMI <30, indicating effects of O3 on lung function in healthy men within the
cohort. However, the findings may be generalizable only to this study population of
older, predominately white men.
Confounding in epidemiologic studies of lung function
The 1996 O3 AQCD noted uncertainty regarding confounding by temperature and
pollen (U.S. EPA. 1996a); however, collective evidence does not indicate that these
factors fully account for the associations observed between increases in ambient O3
concentration and lung function decrements. Across the populations examined, most
studies that found ambient O3-associated lung function decrements, whether
conducted in multiple seasons or only in summer, included temperature in statistical
analyses. Some summer camp studies conducted detailed analysis of temperature.
In most of these studies, temperature and O3 were measured at the camps. In two
6-64
-------
Northeast U.S. studies, an increase in temperature was associated with an increase in
lung function (Thurston et al., 1997; Berry et al., 1991). This positive association
likely accounted for the nearly 2-fold greater decrease in O3-associated PEF found
by Thurston et al. (1997) with temperature in the model than with O3 alone.
In another Northeast U.S. camp study, Spektor et al. (1988a) estimated similar effects
for O3 in a model with and without a temperature-humidity index. In the re-analysis
of six camp studies, investigators did not include temperature in models because
temperature within the normal ambient range had not been shown to affect
O3-induced lung function responses in controlled human exposure studies (Kinnev et
al.. 1996).
Pollen was evaluated in far fewer studies. Camp studies that examined pollen found
that pollen independently was not associated with lung function decrements
(Thurston et al., 1997; Avol et al., 1990). Many studies of individuals with asthma
with follow-up over multiple seasons found O3-associated decrements in lung
function in models that adjusted for pollen counts (Just et al., 2002; Ross et al., 2002;
Jalaludin et al., 2000; Gielen et al., 1997). In these studies, large proportions of
subjects had atopy (22-98%), with some studies examining large proportions of
subjects specifically with pollen allergy who would be more responsive to pollen
exposure (Ross et al., 2002; Gielen et al., 1997).
A relatively large number of studies provided information on potential confounding
by copollutants such as PM2 5, PMi0, NO2, or SO2. While studies were varied in how
they evaluated confounding, most indicated that O3-associated lung function
decrements were not solely due to copollutant confounding. Some studies of subjects
exercising outdoors indicated that ambient concentrations of copollutants such as
NO2, SO2, or acid aerosol were low and thus, not likely to confound associations
observed for O3 (Hoppe et al.. 2003; Brunekreef et al.. 1994; Hoek et al.. 1993).
In other studies of children with increased outdoor exposures, O3 was consistently
associated with decreases in lung function, whereas other pollutants such as PM2 5,
sulfate, and acid aerosol individually showed variable associations across studies
(Thurston et al.. 1997; Castilleios et al.. 1995; Berry et al.. 1991; Avol et al.. 1990;
Spektor et al., 1988a). Most of these studies measured ambient pollutants on site of
subjects' outdoor activity and related lung function changes to the pollutant
concentrations measured during outdoor activity. Thus, the degree of exposure
measurement error likely is comparable for O3 and copollutants.
Studies that conducted copollutant modeling generally found O3-associated lung
function decrements to be robust; i.e., most copollutant-adjusted effect estimates fell
within the 95% CI of the single-pollutant effect estimates (Figure 6-10 [and
Table 6-14]). These studies used central site measurements for both O3 and
copollutants. There may be residual confounding because of differential exposure
measurement error for O3 and copollutants due to differing spatial heterogeneity and
indoor-outdoor ratios; however, the limited available evidence indicates that personal
O3 exposures are weakly correlated with personal PM2 5 and NO2 exposures
(Section 4.3.4.1). Whereas a few studies used the same averaging time for
copollutants (Lewis et al.. 2005; Jalaludin et al.. 2000). more examined 1-h max or
6-65
-------
8-h max O3 and 24-h avg copollutant concentrations (Son et al., 2010; Chen et al.,
1999; Romieu et al., 1997; Romieu et al., 1996). In a Philadelphia-area summer camp
study, Neas et al. (1999) was among the few studies to find that the effect estimate
for O3 was attenuated to near zero in a copollutant model (24-h avg sulfate in this
study) (Figure 6-10 [and Table 6-141).
Ambient O3 concentrations showed a wide range of correlations with copollutant
concentrations (r = -0.31 to 0.74). In Sydney, Australia, Jalaludin et al. (2000) found
low correlations of O3 with PM10 (r = 0.13) and NO2 (r = -0.31), all averaged over
24 hours. In two-pollutant models, PMi0 and NO2 remained associated with
increases in PEF, and O3 remained associated with decreases in PEF in children with
asthma. In Detroit, MI, O3 was moderately correlated with PM2.5 (Pearson r = 0.57)
and PMio (Pearson r = 0.59), all averaged over 24 hours (Lewis et al., 2005).
Adjustment for PMi0 resulted in a large (more negative) change in the O3-associated
FEVi decrement in children with asthma, but only in CS users and not in children
with concurrent URI (Figure 6-10 [and Table 6-141). Studies conducted in Mexico
City, Mexico, found small changes in O3-associated PEF decrements with
copollutant adjustment although different averaging times were used for copollutants
(Romieu et al.. 1997; Romieu et al.. 1996) (Figure 6-10 [Table 6-141). In these
studies, O3 was moderately correlated with copollutants such as NO2 and PMi0
(range of Pearson r = 0.38 - 0.58). Studies conducted in Asia also found that
associations between O3 and lung function were robust to adjustment for weakly- to
moderately-correlated copollutants; effect estimates for copollutants generally were
attenuated, indicating that O3 may confound associations of copollutants (Son et al.,
2010; Chen et al.. 1999).
In a summer camp study conducted in Connecticut, Thurston et al. (1997) found
ambient concentrations of 1-h max O3 and 12-h avg sulfate to be highly correlated
(r = 0.74), making it difficult to separate their independent effects. With sulfate in the
model, a larger decrease in PEF was estimated for O3; however, the 95% CI was
much wider (Figure 6-10 [and Table 6-141). Investigators found that the association
for sulfate was due to one day when the ambient concentrations of both pollutants
were at their peak. With the removal of this peak day, the sulfate effect was
attenuated, whereas O3 effects remained robust (Thurston et al., 1997). Among
children with asthma in Thailand, the O3-associated decrease in PEF was robust to
adjustment of SO2; however, different lags were examined for O3 (lag 5) and SO2
(lag 4) (Wiwatanadate and Trakultivakorn, 2010).
6-66
-------
Study
FEVn
Lewis etal. (2005)
Children with asthma
usingCS
Children with asthma
withURI
Chen etal. (1999)
Children
O3 metrics
24-h avg, Lag 2
24-havg, Lag 2
24-havg, Lag 2
24-h avg, Lag 2
24-h avg, Lag 2
24-h avg, Lag 2
1-h max, Lag 1
1-h max, Lag 1
O3 with copollutant
None •
24-h avg, Lag 1 NO2 O
• k
-15 -13 -11 -9 -7 -5 -3-11 3
Percent change in FEV., per unit increase in O3 (95% Cl)
PEF
Neasetal.(1999)a 12-havg,Lag1
Children attending camp 12-h avg, Lag 1
Thurstonetal. (1997)
Children with asthma
attending camp
Romieu etal. (1996)
Children with asthma
Romieu etal. (1997)
Children with asthma
1-h max, Lag 0
1-h max, Lag 0
1-h max, Lag 0
1-h max, Lag 0
1-h max, Lag 0
1-h max, Lag 0
None
24-h avg, Lag 1 sulfate
None
12-h avg, Lag 0 sulfate
None
24-h avg, Lag 2 PM25
None
24-h avg, Lag 2 PM10
o
-O-
-15 -13 -11 -9 -7 -5 -3-11 3
Percent change in PEF per unit increase in O3 (95% Cl)
Note: Results are presented first for FEV, then for PEF and within these categories, then in order of increasing mean ambient O3
concentration. CS = corticosteroid, URI = Upper respiratory infection. Effect estimates are standardized to a 40-, 30-, and 20-ppb
increase for 1-h max, 12-h avg, and 24-h avg O3, respectively. Black circles represent O3 effect estimates from single pollutant
models, and open circles represent O3 effect estimates from copollutant models.
"Information was not available to calculate 95% Cl of the copollutant model.
Figure 6-10 Comparison of O3-associated changes in lung function in single-
and copollutant models.
6-67
-------
Table 6-14 Comparison of O3-associated changes in lung function in single-
and copollutant models for studies presented in Figure 6-10 plus
others.
Study*
Location/Population
Parameter
Os-associated
Percent Change in
Single-Pollutant
Model (95% Cl)a
O 3-associated
Percent Change in
Copollutant Model
(95% Cl)a
FEV1
Lewis et al.
(2005)
Chen et al.
(1999)
Detroit, Ml
Children with asthma using CS
393 person-days
Children with asthma with URI
231 person-days
Overall mean (SD) age 9.1 (1 .4 yr)
3 Taiwan communities
941 children, ages 8-1 Syr
For 24-h avg, Lag 2
0.29 (-4.2, 5.0)
Lowest daily
FEV,
For 24-h avg, Lag 2
-6.0 (-11. 2, -0.41)
For 1-h max, Lag 1
FEVl -1.5 (-2.8, -0.12)
With 24-h avg, Lag
-0.1 8 (-11. 0,11.9)
With 24-h avg, Lag
-13.4 (-17.8, -8.8)
With 24-h avg, Lag
-5.5 (-10.3, -0.42)
With 24-h avg, Lag
-7.1 (-11.3, -2.8)
With 24-h avg, Lag
-2.0 (-3.5, -0.43)
2 PM2.5
2PM10
2 PM2.5
2PM10
1 NO2
PEF
Neaset al.
(1 999)
Thurston et al.
(1997)
Romieu et al.
(1 996)
Romieu et al.
(1 997)
Philadelphia, PA
1 56 Children at summer camp,
ages 6 - 11 yr
CT River Valley
166 Children with asthma at summer
camp, ages 7-1 Syr
Northern Mexico City, Mexico
71 children with asthma, ages 5-7 yr
Southern Mexico City, Mexico
65 children with asthma, ages 5-1 3 yr
»»««"•*• ŁŁŁŁ'
Intraday For 1-h max, Lag 0
change PEF .2.8 (-4.9, -0.59)
For 1-h max, Lag 2
Evening PEF ^ <_, 3_ „ 19)
For 1-h max, Lag 0
Evening PEF ^52(.10^01)
With 24-h avg, Lag
-0.02b
With 12-h avg, Lag
-11. 8 (-31 .6, 8.1)
With 24-h avg, Lag
-0.24 (-1 .2, 0.68)
With 24-h avg, Lag
-0.79 (-1.4, -0.16)
1 sulfate
0 sulfate
2 PM2.5
OPM10
6-68
-------
Study*
Location/Population
Os-associated
Percent Change in
Single-Pollutant
Parameter Model (95% Cl)a
O 3-associated
Percent Change in
Copollutant Model
(95% Cl)a
Results not included in Figure 6-10°
Jalaludin et al.
(2000)
Sydney, Australia
125 children with asthma or wheeze,
mean (SD) age 9.6 (1 .0) yr
deviation
from mean
PEF
For 15-h (6 a.m.-
9 p.m.) avg, Lag 0
-1.8 (-3.5, -0.19)
With 15-havg, LagOPM10,
-1.8 (-3.5, -0.19)
With 15-havg, Lag 0 NO2
-1.8 (-3.4, -0.11)
Wiwatanadate
and Chiang Mai, Thailand
Daily avg For 24-h avg, Lag 5
PEF (L/min) -2.6 (-5.2, 0)
With 24-h avg, Lag 4 SO2
(2010) " ' '
Son et al.
(2010)
Ulsan, Korea
2,1 02 children and adults, ages 7-97 yr
Change in
percent
predicted
FEV,
For 8-h max, Lag 0-2
avg (kriging)
-1.4 (-2.6, -0.11)
With 24-h avg, Lag2PM10
(kriging)
-1.8 (-3.4, -0.25)
'Includes studies in Figure 6-10 plus others.
CS = Corticosteroid, URI = Upper respiratory infection.
"Effect estimates are standardized to a 40-ppb increase for 1-h max O3, 30-ppb increase for 12-h avg or 8-h max O3, and 20-ppb
increase for 24-h avg or 15-h avg O3.
blnformation was not available to calculate 95% Cl.
°Results are not presented in Figure 6-10 because sufficient data were not available to calculate percent change in lung function.
Some studies did not provide quantitative results but only reported that O3-associated
lung function decrements remained statistically significant in models that included
copollutants such as PMi0, NO2, sulfate, nitrate, or ammonium (Romieu et al..
1998b: Braueretal.. 1996: Linnetal. 1996: Spektor et al.. 1988b).
Several studies estimated robust O3-associated lung function decrements in
multipollutant models that most often included O3, NO2, and either PM2.5 or PM10
(O'Connor et al.. 2008: Thaller et al.. 2008: Chan and Wu. 2005: Romieu et al.. 2002:
Korrick et al.. 1998: Higgins et al.. 1990). However, the independent effects of O3
are more difficult to assess in relation to incremental changes in more than one
copollutant.
Summary of Epidemiologic Studies of Lung Function
The cumulative body of epidemiologic evidence indicates that short-term increases in
ambient O3 concentration are associated with decrements in lung function in children
with asthma (Figure 6-7 [and Table 6-8]) and Figure 6-8 [and Table 6-9]) and
children in the general population. In contrast with results from controlled human
exposure studies, within-study epidemiologic comparisons did not consistently
indicate larger ambient O3-associated lung function decrements in groups with
asthma (children or adults) than in groups without asthma. Notably, most
epidemiologic studies were not designed to assess between-group differences. Based
on comparisons between studies, differences were noted between children with and
without asthma in so far as in studies of children with asthma, an increase in ambient
O3 concentration was associated with both lung function decrements and increases in
6-69
-------
respiratory symptoms (Just et al.. 2002; Mortimer et al.. 2002; Ross et al.. 2002;
Gielenetal. 1997; Romieu etal. 1997; Thurston et al.. 1997; Romieu etal.. 1996).
In studies of children in the general population, increases in ambient O3
concentration were associated with decreases in lung function but not increases in
respiratory symptoms (Ward et al., 2002; Gold etal.. 1999; Linn et al., 1996).
Across studies of children, there was no clear indication that a particular exposure
assessment method using central site measurements produced stronger findings,
despite potential differences in exposure measurement error. In children, lung
function was associated with ambient O3 concentrations measured on site of
children's daytime hours (Hoppe et al., 2003; Thurston et al., 1997), at children's
schools (Chen etal.. 1999; Goldetal.. 1999). at the closest site (Romieu et al.. 2006;
Lewis et al., 2005; Romieu et al., 2004b; Romieu et al., 2002; RomieuetaL, 1997;
Romieu et al., 1996), at multiple community sites then averaged (O'Connor et al.,
2008; Just et al., 2002; Mortimer et al., 2002; Jalaludin et al., 2000), and at a single
site (Ward et al.. 2002; GielenetaL. 1997; Ulmeretal.. 1997; Linn etal.. 1996).
Among children in California, Avol et al. (1998b) found slightly larger O3-associated
lung function decrements for 24-h avg personal exposures than for 1-h max ambient
concentrations.
As noted in the 1996 and 2006 O3 AQCDs, evidence clearly demonstrates ambient
O3-associated lung function decrements in children and adults engaged in outdoor
recreation, exercise, or work. Moreover, several results in these populations indicated
associations with 10-min to 12-h avg O3 concentrations <80 ppb. These studies are
noteworthy for their measurement of ambient O3 on site of and at the time of
subjects' outdoor activity, factors that have contributed to higher O3 personal
exposure-ambient concentration correlations and ratios (Section 4.3.3). These
epidemiologic results are well supported by observations from controlled human
exposure studies of lung function decrements induced by O3 exposure during
exercise (Section 6.2.1.1). Although epidemiologic investigation was relatively
sparse, increases in ambient O3 concentration were not consistently associated with
lung function decrements in adults with respiratory disease, healthy adults, or older
adults.
Across the diverse populations examined, most effect estimates ranged from a <1 to
2% decrease in lung function per unit increase in O3 concentration1. Heterogeneity in
O3-associated respiratory effects within populations was indicated by observations of
larger decreases (3-8%) in children with asthma with CS use or concurrent URI
(Lewis et al.. 2005) and older adults with airway hyperresponsiveness and/or BMI
>30 (Alexeeff et al.. 2007). Among children in Mexico City, Mexico, higher dietary
antioxidant intake attenuated O3-associated lung function decrements (Romieu et al..
2004b; 2002). similar to results from controlled human exposure studies. Each of
these potential effect modifiers was examined in one to two populations; thus, firm
conclusions about their influences are not warranted. Adding to the evidence for
heterogeneity in response, Hoppe et al. (2003) and Mortimer et al. (2002) found that
1 Effect estimates were standardized to a 40-, 30-, and 20-ppb increase for 1-h max, 8-h max, and 24-h avg O3.
6-70
-------
increases in ambient O3 concentration were associated with increased incidence of
>10% decline in lung function in children with asthma.
Collectively, epidemiologic studies examined and found lung function decrements in
association with single-day O3 concentrations lagged from 0 to 7 days and
concentrations averaged over 2-10 days. More studies found associations with O3
concentrations lagged 0 or 1 day (Son et al., 2010; Alexeeff et al., 2008; Lewis et al.,
2005; Ross et al.. 2002; Jalaludin et al.. 2000; Chen et al.. 1999; Romieu et al.. 1997;
Braueretal, 1996; Romieu et al., 1996; Spektor et al., 1988b) than those lagged 5-
7 days (Wiwatanadate and Trakultivakorn, 2010; Hernandez-Cadena et al., 2009;
Steinvil et al., 2009). Associations with multiday average concentrations (Son et al.,
2010; Liu et al.. 2009a; Barraza-Villarreal et al.. 2008; O'Connor et al.. 2008;
Alexeeff et al.. 2007; Mortimer et al.. 2002; Ward et al.. 2002; Goldetal.. 1999;
Naeher et al., 1999; Neas et al., 1999) indicate that elevated exposures over several
days may be important. Within studies, O3 concentrations for multiple lag periods
were associated with lung function decrements, possibly indicating that multiple
modes of action may be involved in the responses. Activation of bronchial C-fibers
(Section 5.3.2) may lead to decreases in lung function as an immediate response to
O3 exposure, and increased airway hyperresponsiveness to antigens resulting from
sensitization of airways by O3 (Section 5.3.5) may mediate lung function responses
associated with lagged or multiday O3 exposures (Peden, 2011).
For single- and multi-day average O3 concentrations, lung function decrements were
associated with 1-h max, 8-h max, and 24-h avg O3, with no strong difference in the
consistency or magnitude of association among the averaging times. For example,
among studies that examined multiple averaging times, Spektor and Lippmann
(1991) found a larger magnitude of association for 1-h max O3 than for 24-h avg O3.
However, other studies found larger magnitudes of association for longer averaging
times [8-h max in Chan and Wu (2005) and 12-h avg in Thaller et al. (2008)1 than for
1-h max O3. Other studies found no clear difference among O3 averaging times
(Jalaludin et al.. 2000; Chen etal.. 1999; Scarlett et al.. 1996; Berry etal.. 1991).
Several studies found that associations with lung function decrements persisted at
lower ambient O3 concentrations. For O3 concentrations averaged up to 1 hour
during outdoor recreation or exercise, associations were found in analyses restricted
to O3 concentrations <80 ppb (Spektor et al.. 1988a; Spektor et al.. 1988b). 60 ppb
(Brunekreefetal.. 1994; Spektor et al.. 1988a). and 50 ppb (Brunekreef etal.. 1994).
Among outdoor workers, Brauer et al. (1996) found a robust association using daily
1-h max O3 concentrations <40 ppb. Ulmer et al. (1997) found a robust association in
schoolchildren using 30-min max O3 concentrations <60 ppb. For 8-h avg O3
concentrations, associations with lung function decrements in children with asthma
were found to persist at concentrations <80 ppb in a U.S. multicity study (for lag 1-5
avg) (Mortimer et al.. 2002) and <51 ppb in a study conducted in the Netherlands
(for lag 2) fGielen et al.. 1997).
Evidence did not demonstrate strong confounding by meteorological factors or
copollutant exposures. Most O3 effect estimates for lung function were robust to
adjustment for temperature, humidity, and copollutants such as PM2.s, PMi0, NO2, or
6-71
-------
SO2. Although examined in few epidemiologic studies, O3 was associated with
decreases in lung function with adjustment for pollen or acid aerosols.
The consistency of association in the collective body of evidence with and without
adjustment for ambient copollutant concentrations and meteorological factors
combined with evidence from controlled human exposure studies for the direct
effects of O3 exposure provide strong support for the independent effects of short-
term ambient O3 exposure on lung function decrements.
6.2.1.3 Toxicology
The 2006 O3 AQCD concluded that pulmonary function decrements occur in a
number of species with acute exposures (< 1 week), ranging from 0.25 to 0.4 ppm O3
(U.S. EPA, 2006b). Early work has demonstrated that during acute exposure of
~0.2 ppm O3 in rats, the most commonly observed alterations are increased
frequency of breathing and decreased tidal volume (i.e., rapid, shallow breathing).
Decreased lung volumes are observed in rats with acute exposures to 0.5 ppm O3.
At concentrations of > 1 ppm, breathing mechanics (compliance and resistance) are
also affected. Exposures of 6 hours/day for 5 days create a pattern of attenuation of
pulmonary function decrements in both rats and humans without concurrent
attenuation of lung injury and morphological changes, indicating that the attenuation
did not result in protection against all the effects of O3 (Tepper et al., 1989).
A number of studies examining the effects of O3 on pulmonary function in rats,
mice, and dogs are described in Table 6-13 on page 6-91 (U.S. EPA, 1996m) of the
1996 O3 AQCD, and Annex Table AX5-11 on page AX5-34 (U.S. EPA. 2006g) of
the 2006 O3 AQCD (U.S. EPA. 2006b. 1996a). Lung imaging studies using
hyperpolarized 3He provide evidence of ventilation abnormalities in rats following
exposure to 0.5 ppm O3 (Cremillieux et al.. 2008). Rats were exposed to 0.5 ppm O3
for 2 or 6 days, either continuously (22 hours/day) or alternatingly (12 hours/day).
Dynamic imaging of lung filling (2 mL/sec) revealed delayed and incomplete filling
of lung segments and lobes. Abnormalities were mainly found in the upper regions of
the lungs and proposed to be due to the spatial distribution of O3 exposure within the
lung. Although the small number of animals used in the study (n = 3 to 7/group)
makes definitive conclusions difficult, the authors suggest that the delayed filling of
lung lobes or segments is likely a result of an increase in airway resistance brought
about by narrowing of the peripheral small airways.
6.2.2 Airway Hyperresponsiveness
Airway hyperresponsiveness refers to a condition in which the conducting airways
undergo enhanced bronchoconstriction in response to a variety of stimuli. Airway
responsiveness is typically quantified by measuring changes in pulmonary function
(e.g., FEVi or specific airway resistance [sRaw]) following the inhalation of an
aerosolized specific (allergen) or nonspecific (e.g., methacholine)
6-72
-------
bronchoconstricting agent or administration of another stimulus such as exercise or
cold air. People with asthma are generally more sensitive to bronchoconstricting
agents than those without asthma, and the use of an airway challenge to inhaled
bronchoconstricting agents is a diagnostic test in asthma. Standards for airway
responsiveness testing have been developed for the clinical laboratory (ATS, 2000a),
although variation in methodology for administering the bronchoconstricting agent
may affect the results (Cockcroft et al.. 2005). There is a wide range of airway
responsiveness in people without asthma, and responsiveness is influenced by a wide
range of factors, including cigarette smoke, pollutant exposures, respiratory
infections, occupational exposures, and respiratory irritants. Airways
hyperresponsiveness in response to O3 exposure has not been examined widely in
epidemiologic studies; such evidence is derived primarily from controlled human
exposure and toxicological studies as described below.
6.2.2.1 Controlled Human Exposures
Beyond its direct effect on lung function, experimental O3 exposure has been shown
to cause an increase in airway responsiveness in human subjects. Increased airway
responsiveness can be an important consequence of exposure to ambient O3, because
the airways are then predisposed to narrowing upon inhalation of a variety of
ambient stimuli.
Increases in airway responsiveness have been reported for exposures to 80 ppb O3
and above. Horstman et al. (1990) evaluated airway responsiveness to methacholine
in young healthy adults (22 M) exposed to 80, 100, and 120 ppb O3 (6.6 hours, quasi
continuous moderate exercise, 39 L/min). Dose-dependent decreases of 33, 47, and
55% in the cumulative dose of methacholine required to produce a 100% increase in
sRaw after exposure to O3 at 80, 100, and 120 ppb, respectively, were reported.
Molfino et al. (1991) reported increased allergen-specific airway responsiveness in
adults with mild asthma exposed to 120 ppb O3 (1 hour resting exposure). Due to
safety concerns, however, the exposures in the Molfino et al. (1991) study were not
randomized with FA conducted first and O3 exposure second. Attempts to reproduce
the findings of Molfino et al. (1991) using a randomized exposure design have not
found statistically significant changes in airway responsiveness at such low levels of
O3 exposure. At a considerably higher exposure to 250 ppb O3 (3 hours, light-to-
moderate intermittent exercise, 30 L/min), Torres et al. (1996) found significant
increases in specific and non-specific airway responsiveness of adults with mild
asthma 3 hours following O3 exposure. Kehrl et al. (1999) found increased reactivity
to house dust mite antigen in adults with mild asthma and atopy 16-18 hours after
exposure to 160 ppb O3 (7.6 hours, light quasi continuous exercise, 25 L/min). Holz
et al. (2002) demonstrated that repeated daily exposure to lower concentrations of
125 ppb O3 (3 hours for four consecutive days; intermittent exercise, 30 L/min)
causes an increased response to allergen challenge at 20 hours postexposure in
allergic airway disease.
6-73
-------
Ozone exposure of subjects with asthma, who characteristically have increased
airway responsiveness at baseline relative to healthy controls (by nearly two orders
of magnitude), can cause further increases in responsiveness (Kreit et al., 1989).
Similar relative changes in airway responsiveness are seen in subjects with asthma
and healthy control subject exposed to O3 despite their markedly different baseline
airway responsiveness. Several studies (Kehrl et al.. 1999: Torres et al.. 1996:
Molfino et al.. 1991) have suggested an increase in specific (i.e., allergen-induced)
airway reactivity. An important aspect of increased airway responsiveness after O3
exposure is that this may provide biological plausibility for associations observed
between increases in ambient O3 concentration and increased respiratory symptoms
in children with asthma (Section 6.2.4.1) and increased hospital admissions and ED
visits for asthma (Section 6.2.7).
Changes in airway responsiveness after O3 exposure appear to resolve more slowly
than changes in FEVi or respiratory symptoms (Folinsbee and Hazucha, 2000).
Studies suggest that O3-induced increases in airway responsiveness usually resolve
18 to 24 hours after exposure, but may persist in some individuals for longer periods
(Folinsbee and Hazucha, 1989). Furthermore, in studies of repeated exposure to O3,
changes in airway responsiveness tend to be somewhat less susceptible to attenuation
with consecutive exposures than changes in FEVi (Gong et al., 1997a: Folinsbee et
al., 1994: Kulle et al., 1982: Dimeo et al., 1981). Increases in airway responsiveness
do not appear to be strongly associated with decrements in lung function or increases
in symptoms (Aris et al., 1995). Recently, Que et al. (2011) assessed methacholine
responsiveness in healthy young adults (83M, 55 F) one day after exposure to
220 ppb O3 and FA for 2.25 hours (alternating 15 min periods of rest and brisk
treadmill walking). Increases in airways responsiveness at 1 day post-O3 exposure
were not correlated with FEVi responses immediately following the O3 exposure or
with changes in epithelial permeability assessed 1 day post-O3 exposure.
6.2.2.2 Toxicology
In addition to human subjects, a number of species, including nonhuman primates,
dogs, cats, rabbits, and rodents, have been used to examine the effect of O3 exposure
on airway hyperresponsiveness (see Table 6-14 on page 6-93 (U.S. EPA, 1996n) of
the 1996 O3 AQCD and Annex Table AX5-12 on page AX5-36 (U.S. EPA. 2006h)
of the 2006 O3 AQCD). With a few exceptions, commonly used animal models have
been guinea pigs, rats, or mice acutely exposed to O3 concentrations of 1 to 3 ppm to
induce airway hyperresponsiveness. These animal models are helpful for determining
underlying mechanisms of general airway hyperresponsiveness and are relevant for
understanding airway responses in humans. Although 1-3 ppm may seem like a high
exposure concentration, based on 18O3 (oxygen-18-labeled O3) in the BALF of
humans and rats, an exposure of 0.4 ppm O3 in exercising humans appears roughly
equivalent to an exposure of 2 ppm in resting rats (Hatch et al., 1994).
6-74
-------
A limited number of studies have observed airway hyperresponsiveness in rodents
and guinea pigs after exposure to less than 0.3 ppm O3. As previously reported in the
2006 O3 AQCD, one study demonstrated that a very low concentration of O3
(0.05 ppm for 4 hours) induced airway hyperresponsiveness in some of the nine
strains of rats tested (Depuydt et al., 1999). This effect occurred at a concentration of
O3 that was much lower than has been reported to induce airway
hyperresponsiveness in any other species. Similar to the effects of O3 on other
endpoints, these observations suggest that a genetic component plays an important
role in O3-induced airway hyperresponsiveness in this species and warrants
verification in other species. More recently, Chhabra et al. (2010) demonstrated that
exposure of ovalbumin (OVA)-sensitized guinea pigs to 0.12 ppm for 2 hours/day for
4 weeks produced specific airway hyperresponsiveness to an inhaled OVA challenge.
Interestingly, in this study, dietary supplementation of the guinea pigs with vitamins
C and E ameliorated a portion of the airway hyperresponsiveness as well as indices
of inflammation and oxidative stress. Larsen and colleagues conducted an O3 C-R
study in mice sensitized by 10 daily inhalation treatments with an OVA aerosol
(Larsen et al.. 2010). Although airway responsiveness to methacholine was increased
in non-sensitized animals exposed to a single 3-hour exposure to 0.5, but not 0.1 or
0.25 ppm O3, airway hyperresponsiveness was observed after exposure to 0.1 and
0.25 ppm O3 in OVA-sensitized mice.
In order to evaluate the ability of O3 to enhance specific and non-specific airway
responsiveness, it is important to take into account the phenomenon of attenuation in
the effects of O3. Several studies have clearly demonstrated that some effects caused
by acute exposure are absent after repeated or prolonged exposures to O3. The ability
of the pulmonary system to adapt to repeated insults to O3 is complex, however, and
experimental findings for attenuation to O3-induced airway hyperresponsiveness are
inconsistent. Airway hyperresponsiveness was observed in mice after a 3-hour
exposure but not in mice exposed continuously for 72 hours to 0.3 ppm (Johnston et
al.. 2005b). However, the Chhabra study demonstrated O3-induced airway
hyperresponsiveness in guinea pigs exposed for 2 hours/day for 10 days (Chhabra et
al.. 2010). Besides the obvious species disparity, these studies differ in that the mice
were exposed continuously for 72 hours, whereas the guinea pigs were exposed
intermittently over 10 days, suggesting that attenuation might be lost with periods of
rest in between O3 exposures.
Overall, numerous toxicological studies have demonstrated that O3-induced airway
hyperresponsiveness occurs in guinea pigs, rats, and mice after either acute or
repeated exposure to relevant concentrations of O3. The mechanisms by which O3
enhances the airway responsiveness to either specific (e.g., OVA) or non-specific
(e.g., methacholine) bronchoprovocation are not clear but appear to be associated
with complex cellular and biochemical changes in the conducting airways. A number
of potential mediators and cells may play a role in O3-induced airway
hyperresponsiveness; mechanistic studies are discussed in greater detail in
Section 5.3.
6-75
-------
6.2.3 Pulmonary Inflammation, Injury and Oxidative Stress
In addition to physiological pulmonary responses, respiratory symptoms, and airway
hyperresponsiveness, O3 exposure has been shown to result in increased epithelial
permeability and respiratory tract inflammation. In general, inflammation can be
considered as the host response to injury and the induction of inflammation as
evidence that injury has occurred. Inflammation induced by exposure of humans to
O3 can have several potential outcomes: (1) inflammation induced by a single
exposure (or several exposures over the course of a summer) can resolve entirely; (2)
continued acute inflammation can evolve into a chronic inflammatory state; (3)
continued inflammation can alter the structure and function of other pulmonary
tissue, leading to diseases such as fibrosis; (4) inflammation can alter the body's host
defense response to inhaled microorganisms, particularly in potentially at-risk
populations such as the very young and old; and (5) inflammation can alter the lung's
response to other agents such as allergens or toxins. Except for outcome (1), the
possible chronic responses have only been directly observed in animals exposed to
O3. It is also possible that the profile of response can be altered in persons with pre-
existing pulmonary disease (e.g., asthma, COPD) or smokers. Oxidative stress has
been shown to play a key role in initiating and sustaining O3-induced inflammation.
Secondary oxidation products formed as a result of reactions between O3 and
components of the ELF can increase the expression of cytokines, chemokines, and
adhesion molecules and enhance airway epithelium permeability (Section 5.3.3. and
Section 5.3.4).
6.2.3.1 Controlled Human Exposures
As reported in studies reviewed in the 1996 and 2006 O3 AQCDs, acute O3 exposure
initiates an acute inflammatory response throughout the respiratory tract that has
been observed to persist for at least 18-24 hours postexposure. A meta-analysis of 21
studies (Mudway and Kelly, 2004a) for varied experimental protocols (80-600 ppb
O3; 1-6.6 hours duration; light to heavy exercise; bronchoscopy at 0-24 hours
post-O3 exposure) showed that neutrophils (PMN) influx in healthy subjects was
linearly associated (p <0.01) with total O3 dose (i.e., the product of O3 concentration,
exposure duration, and VE). As with FEVi responses to O3, within-individual
inflammatory responses to O3 are generally reproducible and correlated between
repeat exposures (Holz et al., 1999). Some individuals also appear to be intrinsically
more susceptible to increased inflammatory responses to O3 exposure (Holz et al.,
2005).
The presence of PMNs in the lung has long been accepted as a hallmark of
inflammation and is an important indicator that O3 causes inflammation in the lungs.
Neutrophilic inflammation of tissues indicates activation of the innate immune
system and requires a complex series of events that are normally followed by
processes that clear the evidence of acute inflammation. Inflammatory effects have
been assessed in vivo by lavage (proximal airway and bronchoalveolar), bronchial
6-76
-------
biopsy, and more recently, induced sputum. A single acute exposure (1-4 hours) of
humans to moderate concentrations of O3 (200-600 ppb) while exercising at
moderate to heavy intensities results in a number of cellular and biochemical changes
in the lung, including an inflammatory response characterized by increased numbers
of PMNs, increased permeability of the epithelial lining of the respiratory tract, cell
damage, and production of proinflammatory cytokines and prostaglandins (U.S.
EPA. 2006b). These changes also occur in humans exposed to 80 and 200 ppb O3 for
6-8 hours (Alexis et al.. 2010: Pedenetal.. 1997: Devlin etal. 1991). Significant
(p = 0.002) increases in sputum PMN (16-18 hours postexposure) relative to FA
responses have been recently reported for 60 ppb O3 which is the lowest exposure
concentration that has been investigated in young healthy adults (Kim et al.. 2011).
Soluble mediators of inflammation such as the cytokines (e.g., IL-6, IL-8) and
arachidonic acid metabolites (e.g., prostaglandin [PG]E2, PGF2a, thromboxane, and
leukotrienes [LTs] such as LTB4) have been measured in the BALF of humans
exposed to O3. In addition to their role in inflammation, many of these compounds
have bronchoconstrictive properties and may be involved in increased airway
responsiveness following O3 exposure. The possible relationship between repetitive
bouts of acute inflammation in humans caused by O3 and the development of chronic
respiratory disease is unknown.
Asthma
Inflammatory responses to O3 exposure have also been studied in subjects with
asthma. Individuals with asthma exposed to 200 ppb O3 for 4-6 hours with exercise
show significantly more neutrophils in BALF (18 hours postexposure) than similarly
exposed healthy individuals (Scannell et al.. 1996: Basha et al.. 1994). In subjects
with allergic asthma who tested positive for Dermatophagoides farinae antigen, there
was an eosinophilic inflammation (2-fold increase), as well as neutrophilic
inflammation (3-fold increase) 18 hours after exposure to 160 ppb O3 for 7.6 hours
with exercise (Peden et al.. 1997). In a study of subjects with intermittent asthma
exposed to 400 ppb O3 for 2 hours, increases in eosinophil cationic protein,
neutrophil elastase and IL-8 were found to be significantly increased 16 hours
postexposure and comparable in induced sputum and BALF (Hiltermann et al..
1999). At 18 hours post-O3 exposure (200 ppb, 4 hours with exercise) and corrected
for FA responses, Scannell et al. (1996) found significantly increased neutrophils in
18 adults with asthma (12%) compared to 20 healthy subjects (4.5%). This difference
in inflammatory response was observed despite no group differences in spirometric
responses to O3. Scannell et al. (1996) also reported that IL-8 tends to be higher in
the BALF of subjects with asthma compared to those without asthma following O3
exposure, suggesting a possible mediator for the significantly increased neutrophilic
inflammation in those subjects. Bosson et al. (2003) found significantly greater
epithelial expression of IL-5, IL-8, granulocyte-macrophage colony-stimulating
factor and epithelial cell-derived neutrophil-activating peptide-78 in subjects with
asthma compared to healthy subjects following exposure to 200 ppb O3 for 2 hours.
In contrast, Stenfors et al. (2002) did not detect a difference in the O3-induced
increases in neutrophil numbers between 15 subjects with mild asthma and 15
6-77
-------
healthy subjects by bronchial wash at the 6 hours postexposure time point. However,
the subjects with asthma were on average 5 years older than the healthy subjects in
this study, and it is not yet known how age affects inflammatory responses. It is also
possible that the time course of neutrophil influx differs between healthy individuals
and those with asthma. Differences between subjects with asthma and healthy
subjects in O3-mediated activation of innate and adaptive immune responses have
been observed in two studies (Hernandez et al.. 2010: Bosson et al.. 2003). as
discussed in Section 6.2.5.4 and Section 5.4.2.2.
Vagaggini et al. (2002) investigated the effect of prior allergen challenge on
responses in subjects with mild asthma exposed for 2 hours to 270 ppb O3 or filtered
air. At 6 hours postexposure, eosinophil numbers in induced sputum were found to
be significantly greater after O3 than after air exposures. Studies such as this suggest
that the time course of eosinophil and neutrophil influx following O3 exposure can
occur at levels detectable within the airway lumen by as early as 6 hours. They also
suggest that the previous or concurrent activation of pro-inflammatory pathways
within the airway epithelium may enhance the inflammatory effects of O3. For
example, in an in vitro study of primary human nasal epithelial cells and BEAS-2B
cell line, cytokine production induced by rhinovirus infection was enhanced
synergistically by concurrent exposure to O3 at 200 ppb for 3 hours (Spannhake et
al.. 2002).
A few studies have evaluated the effects of corticosteroid usage on the response of
subjects with asthma to O3. Vagaggini et al. (2007) evaluated whether corticosteroid
usage would prevent O3-induced lung function decrements and inflammatory
responses in a group of subjects with mild persistent asthma (n = 9; 25 ± 7 years).
In this study, subjects with asthma were randomly exposed on four occasions to
270 ppb O3 or FA for 2 hours with moderate exercise. Exposures were preceded by
four days of treatment with prednisone or placebo. Pretreatment with corticosteroids
prevented an inflammatory response in induced sputum at 6 hours postexposure.
FEVi responses were, however, not prevented by corticosteroid treatment and were
roughly equivalent to those observed following placebo. Vagaggini et al. (2001) also
found budesonide to decrease airway neutrophil influx in subj ects with asthma
following O3 exposure. In contrast, inhalation of corticosteroid budesonide failed to
prevent or attenuate O3-induced responses in healthy subjects as assessed by
measurements of lung function, bronchial reactivity and airway inflammation
(Nightingale et al., 2000). High doses of inhaled fluticasone and oral prednisolone
have each been reported to reduce inflammatory responses to O3 in healthy
individuals (Holz et al.. 2005).
Stenfors et al. (2010) exposed adults with persistent asthma (n = 13; aged 33 years)
receiving chronic inhaled corticosteroid therapy to 200 ppb O3 for 2 hours with
moderate exercise. At 18 hours postexposure, there was a significant O3-induced
increase in bronchioalveolar lavage (BAL) neutrophils, but not eosinophils.
Bronchial biopsy also showed a significant O3-induced increase in mast cells.
Results from this study suggest that the protective effect of acute corticosteroid
6-78
-------
therapy against inflammatory responses to O3 in subjects with asthma demonstrated
by Vagaggini et al. (2007) may be lost with continued treatment regimes.
Associations between Inflammation and FEV1 responses
Studies reviewed in the 2006 O3 AQCD reported that inflammatory responses do not
appear to be correlated with lung function responses in either subjects with asthma or
healthy subjects. In healthy adults (14 M, 6 F) and volunteers with asthma (12 M, 6
F) exposed to 200 ppb O3 (4 hours with moderate quasi continuous exercise, VE = 44
L/min), percent PMN and total protein in BAL fluids were significantly increased in
the subjects with asthma relative to the healthy controls. Spirometric measures of
lung function were significantly decreased following the O3 exposure in both groups,
but were not significantly different between the subjects with asthma and healthy
subjects. Effects of O3 on PMN and total protein were not correlated with changes in
FEVi or FVC (Balmes et al.. 1997: Balmes et al.. 1996). Devlin et al. (1991) exposed
healthy adults (18 M) to 80 and 100 ppb O3 (6.6-hours with moderate quasi
continuous exercise, 40 L/min). In BAL fluid collected 18 hours after exposure to
100 ppb O3, significant increases in PMNs, protein, PGE2, fibronectin, IL-6, lactate
dehydrogenase, and a-1 antitrypsin were found compared to FA. Similar but smaller
increases in all mediators were found after exposure to 80 ppb O3 except for protein
and fibronectin. Changes in BAL markers were not correlated with changes in FEVi.
Holz et al. (1999) examined inflammatory responses in healthy subjects (n = 21) and
those with asthma (n = 15) exposed to 125 and 250 ppb O3 (3 hours, light
intermittent exercise, 26 L/min). Significantly increased percent PMN in sputum due
to O3 exposure was observed in both healthy subjects and those with asthma
following the 250 ppb exposure. At the lower 125 ppb exposure, only the group with
asthma experienced statistically significant increases in the percent PMN. Significant
decrements in FEVi were only found following exposure to 250 ppb; these changes
in FEVi did not differ significantly between the group with asthma and healthy
group and were not correlated with changes in PMN levels. Peden et al. (1997) also
found no correlation between PMN and FEVi responses in 8 individuals with asthma
exposed to 160 ppb O3 for 7.6 hours with light-to-moderate exercise (VE = 25
L/min). However, a marginally significant correlation (r = -0.69, two-tailed p = 0.08,
n = 7) was observed between increases in the percentage of eosinophils and FEVi
responses following O3 exposure.
In contrast to these earlier findings, Vagaggini et al. (2010) recently reported a
significant (r = 0.61, p = 0.015) correlation between changes in FEVi and changes in
sputum neutrophils in subjects with mild-to-moderate asthma (n = 23; 33 ± 11 years)
exposed to 300 ppb O3 for 2 hours with moderate exercise. Eight subjects were
categorized as "responders" based on >10% FEVi decrements. There were no
baseline differences between responders and nonresponders. However, at 6 hours
post-O3 exposure, sputum neutrophils were significantly increased by 15% relative to
FA in responders. The neutrophil increase in responders was also significantly
greater than the 0.2% increase in nonresponders. Interestingly, the nonresponders in
the Vagaggini et al. (2010) study experienced a significant O3-induced 11.3%
6-79
-------
increase in sputum eosinophils, while responders had an nonsignificant 2.6%
decrease.
Time Course of the Inflammatory Response
The time course of the inflammatory response to O3 in humans has not been fully
characterized. Different markers exhibit peak responses at different times. Studies in
which lavages were performed 1 hour after O3 exposure (1 hours at 400 ppb or 4
hours at 200 ppb) have demonstrated that the inflammatory responses are quickly
initiated (Torres et al.. 1997: Devlin etal.. 1996: Schelegle et al.. 1991).
Inflammatory mediators and cytokines such as IL-8, IL-6, and PGE2 are greater at
1 hour than at 18 hours post-O3 exposure (Torres et al., 1997: Devlin et al., 1996).
However, IL-8 still remained elevated at 18 hours post-O3 exposure (4 hours at
200 ppb O3 versus FA) in healthy subjects (Balmes et al., 1996). Schelegle et al.
(1991) found increased PMNs in the "proximal airway" lavage at 1, 6, and 24 hours
after O3 exposure (4 hours at 200 ppb O3), with a peak response at 6 hours.
However, at 18-24 hours after O3 exposure, PMNs remain elevated relative to 1 hour
postexposure (Torres et al., 1997: Schelegle et al., 1991).
Genetic Polymorphisms
Alexis et al. (2010) recently reported that a 6.6-hour exposure with moderate exercise
to 80 ppb O3 caused increased sputum neutrophil levels at 18 hours postexposure in
young healthy adults (n = 15; 24 ± 1 years). In a prior study, Alexis et al. (2009)
found genotype effects on inflammatory responses to O3, but not lung function
responses following a 2-hour exposure to 400 ppb O3. At 4 hours post-O3 exposure,
groups of both GSTM1 genotypes had significant increases in sputum neutrophils
with a tendency for a greater increase in GSTM1-sufficient than null individuals.
At 24 hours postexposure, neutrophils had returned to baseline levels in the GSTM1 -
sufficient individuals. In the GSTMl-null subjects, however, neutrophil levels
increased further from 4 hours to 24 hours and were significantly greater than both
baseline levels and 24 hours levels in GSTM1-sufficient individuals. Alexis et al.
(2009) found that GSTM1-sufficient individuals (n = 19; 24 ± 3 years) had a
decrease in macrophage levels at 4-24 hours postexposure to 400 ppb O3 for 2 hours
with exercise. These studies also provide evidence for activation of innate immunity
and antigen presentation, as discussed in Section 5.3.6. Effects of the exposure apart
from O3 cannot be ruled out in the Alexis et al. (2010): (2009) studies, however,
since no FA exposure was conducted.
Vagaggini et al. (2010) examined FEVi and sputum neutrophils in subjects with
mild-to-moderate asthma (n = 23; 33 ± 11 years) exposed to 300 ppb O3 for 2 hours
with moderate exercise. Six of the subjects were NQO1 wild type and GSTM1 null,
but this genotype was not found to be associated with O3-induced changes in lung
function or inflammatory responses to O3. Kim et al. (2011) showed a significant
(p = 0.002) increase in sputum neutrophil levels following a 6.6-hour exposure to
6-80
-------
60 ppb O3 relative to FA in young healthy adults (13 F, 11 M; 25.0 ±0.5 years).
There was no significant effect of GSTM1 genotype (half GSTM1-null) on the
inflammatory responses observed in these individuals. Previously, inflammatory
responses had only been evaluated down to a level of 80 ppb O3.
Repeated Exposures
Changes in markers from BALF following both 2-hour (Devlin et al., 1997) and 4-
hour (Jorres et al., 2000; Christian et al., 1998) repeated O3 exposures (up to 5 days)
indicate that there is ongoing cellular damage irrespective of the attenuation of some
cellular inflammatory responses of the airways, pulmonary function, and symptom
responses. Devlin et al. (1997) found that several indicators of inflammation
(e.g., PMN, IL-6, PGE2, fibronectin) were attenuated after 5 days of exposure
(i.e., values were not different from FA). However, other markers (LDH, IL-8, total
protein, epithelial cells) did not show attenuation, suggesting that tissue damage
probably continues to occur during repeated exposure. Some cellular responses did
not return to baseline levels for more than 10-20 days following O3 exposure.
Christian et al. (1998) showed decreased numbers of neutrophils and a decrease in
IL-6 levels in healthy adults after 4 days of exposure versus the single exposure to
200 ppb O3 for 4 hours. Jorres et al. (2000) also found that both functional and
BALF cellular responses to O3 were abolished at 24 hours postexposure following
the fourth exposure day. However, levels of total protein, IL-6, IL-8, reduced
glutathione and ortho-tyrosine were still increased significantly. In addition, visual
scores (bronchoscopy) for bronchitis and erythema and the numbers of neutrophils in
bronchial mucosal biopsies were increased. Results indicate that, despite an
attenuation of some markers of inflammation in BALF and pulmonary function
decrements, inflammation within the airways persists following repeated exposure to
O3. The continued presence of cellular injury markers indicates a persistent effect
that may not necessarily be recognized due to the attenuation of spirometric and
symptom responses.
Epithelial Permeability
A number of studies show that O3 exposures increase epithelial cell permeability
through direct (technetium-99m labeled diethylene triamine pentaacetic acid,
99mTc-DTPA, clearance) and indirect (e.g., increased BALF albumin, protein)
techniques. Kehrl et al. (1987) showed increased 99mTc-DTPA clearance in healthy
young adults (age 20-30 yrs) at 75 minutes postexposure to 400 ppb O3 for 2 hours.
Also in healthy young adults (age 26 ± 2 yrs), Foster and Stetkiewicz (1996) have
3
shown that increased mTc-DTPA clearance persists for at least 18-20 hours post-O
exposure (130 minutes to average O3 concentration of 240 ppb), and the effect is
greater at the lung apices than at the base. In a older group of healthy adults (mean
age = 35 yrs), Morrison et al. (2006) observed 99mTc-DTPA clearance at 1 hours and
6 hours postexposure to O3 (100 and 400 ppb; 1 hour; moderate intermittent exercise,
6-81
-------
VE = 40 L/min) to be similar and not statistically different from 99mTc-DTPA
clearance at 1-hour postexposure to FA (1 hour; VE = 40 L/min).
Increased BALF protein, suggesting O3-induced changes in epithelial permeability,
have also been reported at 1 hour and 18 hours postexposure (Devlin et al.. 1997:
Balmes et al.. 1996). Meta-analysis of results from 21 publications (Mudwav and
Kelly. 2004a) for varied experimental protocols (80-600 ppb O3; 1-6.6 hours
duration; light to heavy exercise; bronchoscopy at 0-24 hours post-O3 exposure),
showed that increased BALF protein is associated with total inhaled O3 dose (i.e., the
product of O3 concentration, exposure duration, and VE).
It has been postulated that changes in permeability associated with acute
inflammation may provide increased access of inhaled antigens, particles, and other
inhaled substances deposited on lung surfaces to the smooth muscle, interstitial cells,
and the blood. Hence, increases in epithelial permeability following O3 exposure
might lead to increases in airway responsiveness to specific and nonspecific agents.
Que et al. (2011) investigated this hypothesis in healthy young adults (83M, 55 F)
exposed to 220 ppb O3 for 2.25 hours (alternating 15 min periods of rest and brisk
treadmill walking). As has been observed by others for FEVi responses, within-
individual changes in permeability were correlated between sequential O3 exposures.
This indicates intrinsic differences in susceptibility to epithelial damage from O3
exposure among individuals. Increases in epithelial permeability at 1 day post-O3
exposure were not correlated with FEVi responses immediately following O3
exposure or with changes in airway responsiveness to methacholine assessed 1 day
post-O3 exposure. The authors concluded that changes in FEVi, permeability, and
airway responsiveness following O3 exposure were relatively constant over time in
young healthy adults, although these responses appear to be mediated by differing
physiologic pathways.
6.2.3.2 Epidemiology
In the 2006 O3 AQCD, epidemiologic evidence of associations between short-term
increases in ambient O3 concentration (30-min or 1-h max) and changes in
pulmonary inflammation was limited to a few observations of increases in nasal
lavage levels of inflammatory cell counts, eosinophilic cationic protein, and
myeloperoxidases (U.S. EPA, 2006b). In recent years, there has been a large increase
in the number of studies assessing ambient O3-related changes in pulmonary
inflammation and oxidative stress, types of biological samples collected (i.e., lower
airway), and types of indicators examined. Most studies collected samples every 1 to
3 weeks resulting in a total of 3 to 8 samples per subject. These recent studies form a
larger base to establish coherence with findings from controlled human exposure and
animal studies that have measured the same or related biological markers.
Additionally, results from these studies provide further biological plausibility for the
associations observed between ambient O3 concentrations and respiratory symptoms
and asthma exacerbations.
6-82
-------
Despite the strengths of studies of inflammation, research in this field continues to
develop, and several uncertainties are recognized that may limit inferences from
results indicating the effects of ambient O3 exposure. Current areas of development
include examination of the clinical relevance of the observed magnitudes of changes
in biological markers of pulmonary inflammation (Murugan et al.. 2009; Duramad et
al.. 2007). characterization of the time course of changes between biomarker levels
and other endpoints of respiratory morbidity, development of standardized
methodologies for collection, improvement of the sensitivity and specificity of assay
methods, and characterization of subject factors (e.g., asthma severity, recent
medication use) that contribute to inter-individual variability. These sources of
uncertainty may contribute to differences in findings among studies.
Although most of the biomarkers examined in epidemiologic studies were not
specific to the lung, most studies collected exhaled breath, exhaled breath condensate
(EEC), nasal lavage fluid, or induced sputum with the aim of monitoring
inflammatory responses in airways, as opposed to monitoring systemic responses in
blood. The biomarker most frequently measured was exhaled nitric oxide (eNO),
likely related to its ease of collection in the field and automated measurement. Other
biological markers were examined in EEC, induced sputum, and nasal lavage fluid,
which are hypothesized to represent the fluid lining the lower or upper airways and
contain cytokines, cells, and/or markers of oxidative stress that mediate inflammatory
responses (Balbi et al.. 2007: Howarth et al.. 2005: Hunt 2002). Table 6-15 presents
the locations, time periods, and ambient O3 concentrations for studies examining
associations with biological markers of pulmonary inflammation and oxi dative stress.
Many studies found that short-term increases in ambient O3 concentration were
associated with increases in pulmonary inflammation and oxi dative stress, in
particular, studies of children with asthma conducted in Mexico City, Mexico
(Figure 6-11 [and Table 6-161 and Table 6-17).
6-83
-------
Table 6-15 Mean and upper percentile O3 concentrations in epidemiologic
studies of biological markers of pulmonary inflammation and
oxidative stress.
Study*
Barraza-
Villarreal et
al. (2008)
Berhane et
al. (2011)
Liu et al.
(2009a)
Khatri et al.
(2009)
Qian et al.
(2009)
Romieu et al.
(2008)
Sienra-
Monge et al.
(2004)
Ferdinands
et al. (2008)
Chimenti et
al. (2009)
Nickmilderet
al. (2007)
Delfino et al.
(201 Oa)
Adamkiewicz
et al. (2004)
Location
Mexico City, Mexico
13 Southern California
Communities
Windsor, ON, Canada
Atlanta, GA
Boston, MA; New
York, NY; Denver,
CO; Philadelphia, PA;
San Francisco, CA;
Madison, Wl (SOCS)
Mexico City, Mexico
Mexico City, Mexico
Suburb of Atlanta, GA
Sicily, Italy
Southern Belgium
Los Angeles, CA
Steubenville, OH
Study
Period
June 2003-
June 2005
Sept 2004-
June 2005
Oct-Dec
2005
May-Sept
2003, 2005,
2006
Feb 1997-
Jan 1999
Jan-Oct
2004
All-year
1 999-2000
Aug 2004
Nov, Feb,
July,
year NR
July-Aug
2002
Warm and
cold season
2005-2007
Sept-Dec
2000
O3 Averaging
Time
8-h moving avg
8-h avg
(10a.m.-
6 p.m.)
24-h avg
8-h max
8-h max
8-h max
8-h max
1 -h max
8-h avg
(7 a.m. -3 p.m.)
1 -h max
8-h max
24-h avg
24-h avg
1 -h avgd
Mean/ Median
Concentration (ppb)
31.6
NR
13.0
median: 61 a
33.6
31.1
66.2
61 (median)
November: 32.7 (pre-race),
35.1 (race)"
February: 37.0 (pre-race),
30.8 (race)"
July: 51 .2 (pre-race),
46.1 (race)"
NR
NR
Warm season median:
32.1°
Cool season median: 19.1°
15.3
19.8
Upper Percentile
Concentrations (ppb)
Max: 86.3
NR
95th: 26.5
75th: 74a
75th: 44.4, Max: 91 .5
75th: 38.3, Max: 60.7
Max: 142.5
75th: 67
NR
Max (across 6 camps):
24.6-112.8"
Max (across 6 camps):
19.0-81.1"
Max: 76.4°
Max: 44.9°
75th: 20.2, Max: 32.2
75th: 27.5, Max: 61 .6
*Note: Studies presented in order of first appearance in the text of this section.
NR = Not Reported, SOCS = Salmeterol Off Corticosteroids Study.
alndividual-level estimates were calculated based on time spent in the vicinity of various O3 monitors.
"Concentrations converted from ug/m3 to ppb using the conversion factor of 0.51 assuming standard temperature (25°C) and
pressure (1 atm).
°Measurements outside subject's residence (retirement home).
dAverage O3 concentration in the 1 hour preceding eNO collection.
6-84
-------
Study O3 Lag Subgroup
Individuals with asthma
i in at ^i C3r\r\o\ n ^ A
LIU 61 al. (Ł\j\Jo) U ^ V
Barraza-Villarreal et _ ....... , ,.
al.(2008) ° Without asthma
Children With asthma
Berhane et al. 1-23 cum avg Without asthma
(2011) \A/ith asthma
Children
Without allergy
With allergy
Qian et al. (2009) 0 •
Children and adults _ 0 m
with asthma °-3av9 *
Khatrietal. (2009) 2
Adults with asthma
Older adults
Adamkiewiczetal. 0, 24-h avg _
(2004)
01-h nvn ^ A-
Delfinoetal. (2010) 0-4 avg Cool season
• b-
• b-
• b-
• b-
-30 -20 -10 0 10 20 30 40 50
Percent change in eNO per unit increase in O3(95% Cl)
Note: Results are presented first for children with asthma and adults with asthma then for older adults. Effect estimates are from
single-pollutant models and are standardized to a 40-ppb increase for 1-h avg O3 concentrations, a 30-ppb increase for 8-h max
or 8-h avg O3 concentrations, and a 20-ppb increase for 24-h avg O3 concentrations.
Figure 6-11 Percent change in exhaled nitric oxide (eNO) in association with
ambient O3 concentrations in populations with and without asthma.
6-85
-------
Table 6-16 Percent change in exhaled nitric oxide (eNO) in association with
ambient O3 concentrations in populations with and without asthma
for studies presented in Figure 6-11.
Study*
Location/Population
03
Averaging
Time
O3 Lag
Standardized
% Change
Subgroup (95% Cl)a
Studies in individuals with asthma
Liu et al. (2009a)
Windsor, ON, Canada
182 children with asthma, ages 9-
14yr
Barraza-Villarreal et Mexico City, Mexico
al. (2008) 208 children, ages 6-1 4 yr
Berhane et al.
(2011)
Qian et al. (2009)
Khatri et al. (2009)
Studies in older
Adamkiewicz et al.
(2004)
Delfino et al.
(201 Oa)
13 Southern California communities
2,240 children, ages 6-10 yr
Boston, MA; New York, NY;
Denver, CO; Philadelphia, PA; San
Francisco, CA; Madison, Wl
1 1 9 children and adults with
asthma, ages 12-65yr
Atlanta, GA
38 adults with asthma, ages 31 -
50 yr
adults
Steubenville, Ohio
29 older adults, ages 53-90 yr
Los Angeles, CA
60 older adults, ages a 65 yr
24-h avg
8-h moving avg
8-h avg
(10a.m.-
6 p.m.)
8-h max
8-h max
24-h avg
1 -h avgb
24-h avg
0
1
0
1-23
cumulative
avg
0
0-3 avg
2
0
0-4 avg
-25.1 (-42.9, -1.7)
-17.5 (-32.1, -0.24)
Without asthma 13.5(11 .2, 1 5.8)
With asthma 6.2 (6.0, 6.5)
Without asthma 30.1 (10.6, 53.2)
With asthma 26.0 (-1 .4, 60.9)
Without allergy 25.5 (5.3, 49.6)
With allergy 32.1 (12.0, 55.9)
-1.2 (-1.7, -0.64)
-1.0 (-1.8, -0.26)
36.6(1.2,72.0)
-5.7 (-25.9, 14.5)
-19.7 (-41 .3, 1.9)
Cool season 35.2(10.9,59.5)
Warm season -0.60 (-1 4.0, 1 2.8)
'Includes studies in Figure 6-11.
"Effect estimates are standardized to a 40-ppb, 30-ppb, and 20-ppb increase
respectively.
bAverage O3 concentration in the 1 hour preceding eNO collection.
for 1 -h avg, 8-h max or 8-h avg, and 24-h avg O3,
6-86
-------
Table 6-17 Associations between short-term ambient O3 exposure and
biological markers of pulmonary inflammation and oxidative stress.
Study
Liu et al.
(2009a)
Romieu et al.
(2008)
Barraza-
Villarreal et
al. (2008)
Sienra-Monge
et al. (2004)
Khatri et al.
(2009)
Ferdinands et
al. (2008)
Location/Population
Windsor, ON, Canada
182 children with asthma,
ages 9 - 14yr
Mexico City, Mexico
107 children with asthma,
mean (SD)
age 9.5 (2.1) yr
Mexico City, Mexico
208 children,
ages 6 - 14yr
Mexico City, Mexico
1 17 children with asthma,
mean age 9 yr
Atlanta, GA
38 adults with asthma,
ages 31 - 50 yr
Atlanta, GA
16 children exercising
outdoors, ages 1 4 - 1 7 yr
03
Averaging O3 Biological
Time Lag Marker
EEC
8-isoprostane
24-h ava 0 <% Chan9e>
EEC TEARS
(% change)
EEC
8-h max 0 Ma|ondia|dehydeb
Nasal lavage IL-8
(pg/mL)
8-h moving „
avg
EBCpH
Nasal lavage
IL-8b
Nasal lavage
0-2 IL-6b
0 u m-,v U Z
"~h avg
Nasal lavage
Uricacidb
Nasal lavage
Glutathione11
D , 0 Blood eosinophils
8-h max 2 (OX) change)
EBCpH
1-hmax 0 (normalized
score)
Subgroup
Without
asthma
With asthma
Without
asthma
With asthma
Placebo
Antioxidant
Placebo
Antioxidant
Placebo
Antioxidant
Placebo
Antioxidant
Standardized
Effect Estimate
(95% Cl)a
16.2 (-13.9, 56.8)
11. 5 (-27.0, 70.1)
1.9(1.1,3.5)
1.6(1.4, 1.9)
1.6(1.4, 1.8)
-0.10 (-0.27, 0.08)°
-0.10 (-0.20, 0.01)°
2.2(1.1,4.7)
1.0(0.44, 2.3)
2.7(1.4,5.1)
1.1 (0.53, 2.2)
0.75(0.44, 1.3)
1.3(0.68,2.4)
0.79 (0.63, 0.98)
0.80 (0.66, 0.96)
2.4(0.62,4.2)
0.80 (-0.20, 1 .8)°
Results generally are presented in order of increasing mean ambient O3 concentration. EBC = exhaled breath condensate,
TEARS = thiobarbituric acid reactive substances, IL-8 = interleukin 8, IL-6 = interleukin 6, Antioxidant = group supplemented with
vitamins C and E.
"Effect estimates are standardized to a 40-, 30- and 20-ppb increase for 1 -h max, 8-h max or 8-h avg, and 24-h avg O3,
respectively.
bEffect estimates represent the ratio of the geometric means of biological marker per unit increase in O3 concentration. A ratio <1
indicates a decrease in marker, and a ratio >1 indicates an increase in marker for an increase in O3.
°Model analyzed log-transformed O3. Decreases and increases in pH indicate increases and decreases in pulmonary inflammation,
respectively.
6-87
-------
Populations with Asthma
Exhaled Nitric Oxide
Neither NO nor eNO has been examined in the controlled human exposure or
toxicological studies of O3 exposure reviewed in this ISA. However, several lines of
evidence support its analysis as an indicator of pulmonary inflammation. Inducible
NO synthase can be activated by pro-inflammatory cytokines, and NO can be
produced by cells such as neutrophils, eosinophils, and epithelial cells in the lung
during an inflammatory response (Barnes andLiew. 1995). Further, eNO commonly
is higher in individuals with asthma and increases during acute exacerbations (Jones
et al.. 2001: Kharitonov and Barnes. 2000).
As indicated in Figure 6-10 [and Table 6-16]). short-term increases in ambient O3
concentration (8-h max or avg) were associated with increases in eNO in children
with asthma. These studies used different methods to assign exposures using central
site O3 measurements: the site closest (within 5 km) to home or school (Barraza-
Villarreal et al., 2008) and a single site per community (Berhane et al., 2011).
Because information on spatial homogeneity of ambient O3 concentrations and time
spent outdoors was not available in these studies, it is not possible to assess whether
these two methods produced different personal-ambient O3 ratios and correlations.
Liu et al. (2009a) (described in Section 6.2.1.2) reported O3-associated decreases in
eNO; however, this study was restricted to winter. Results for EBC markers of
oxidative stress and lung function collectively also provided weak evidence of
O3-associated respiratory effects in this study. As described in Section 4.3.3, in non-
summer months, indoor to outdoor O3 ratios are lower as are personal-ambient ratios,
making it more difficult to detect associations with ambient O3 concentrations.
In contrast with controlled human exposure studies (Section 6.2.3.1). epidemiologic
studies did not find larger O3-associated increases in pulmonary inflammation in
groups with asthma than in groups without asthma (Figure 6-11 [and Table 6-16]).
Among children in Southern California, Berhane et al. (2011) estimated similar
associations for a 1-23 day cumulative average of 8-h avg (10 a.m.-6 p.m.) O3 in
children with and without asthma. Among children in Mexico City, Mexico, Barraza-
Villarreal et al. (2008) found a larger association (for lag 0 [of 8-max O3]) in
children without asthma, most of whom had atopy.
Studies that included adults with asthma produced contrasting results (Khatri et al..
2009: Qian et al.. 2009). The multicity salmeterol ((3-2 agonist) trial (Boston, MA;
New York, NY; Denver, CO; Philadelphia, PA; San Francisco, CA; and Madison,
WI) involved serial collection of eNO from 119 subjects with asthma, 87% of whom
were 20-65 years of age (Qian et al.. 2009). Ambient O3 concentrations were
averaged from all sites within 20 miles of subjects' zipcode centroids, which in a
repeated measures study, may capture the temporal variation in O3 reasonably well
(Darrow et al.. 201 la: Gent et al.. 2003). Among all subjects, increases in 8-h max
O3 at multiple lags (0 to 3 single-day and 0-4 avg) were associated with decreases in
eNO. Results did not vary among the salmeterol-, CS-, and placebo-treated groups,
indicating that the counterintuitive findings for O3 were not only due to the reduction
-------
of inflammation by medication. Qian et al. (2009) suggested that at higher
concentrations, O3 may transform NO in airways to reactive nitrogen species.
However, this hypothesis was not supported by results from Khatri et al. (2009), who
in Atlanta, GA examined overall higher 8-h max O3 ambient concentrations than did
Qian et al. (2009) and found that an increase in O3 was associated with an increase in
eNO in adults with asthma (36.6% [95% CI: 1.2, 71.9] per 30-ppb increase in lag 2
of 8-h max O3). Although Khatri et al. (2009) was cross-sectional and did not adjust
for any meteorological factors, it may have better characterized O3 exposures
because subjects were examined during warm months, and an 8-h max O3
concentration was calculated for each subject using measurements at the site closest
to his/her location each hour.
Other biological markers of pulmonary inflammation and oxidative
stress
Short-term increases in ambient O3 concentration were associated with changes in
the levels of pro-inflammatory cytokines and cells, indicators of oxidative stress, and
antioxidants (Table 6-17). Importantly, any particular biomarker was examined in
only one to two studies, and the evidence in individuals with asthma is derived
primarily from studies conducted in Mexico City, Mexico (Barraza-Villarreal et al..
2008: Romieu et al.. 2008: Sienra-Monge et al.. 2004). These studies measured
ambient O3 concentrations at sites within 5 km of subjects' schools or homes. In a
Mexico City cohort of children with asthma, school ambient O3 concentrations
averaged over 48 to 72 hours had a ratio and correlation with personal exposures (48-
to 72-h avg) of 0.17 and 0.35, respectively (Ramirez-Aguilar et al.. 2008). These
observations suggest that the effects of personal O3 exposure on inflammation may
have been underestimated in the Mexico City studies. Despite the limited evidence,
the epidemiologic findings are well supported by controlled human exposure and
toxicological studies that measured the same or related endpoints.
Several of the modes of action of O3 are mediated by reactive oxygen species (ROS)
produced in the airways by O3 (Section 5.3.3). These ROS are important mediators
of inflammation as they regulate cytokine expression and inflammatory cell activity
in airways (Heidenfelder et al., 2009). Controlled human exposure and toxicological
studies, frequently have found O3-induced increases in oxidative stress as shown by
increases in prostaglandins (Section 5.3.3 and Section 6.2.3.1), which are produced
by the peroxidation of cell membrane phospholipids (Morrow et al.. 1990). Romieu
et al. (2008) analyzed EEC malondialdehyde (MDA), a thiobarbituric acid reactive
substance, which like prostaglandins, is derived from lipid peroxidation (Janero.
1990). For a 30-ppb increase in lag 0 of 8-h max O3, the ratio of the geometric means
of MDA was 1.9 (95% CI: 1.1,3.5). Similar results were reported for lags 1 and 0-1
avg O3. A limitation of the study was that 25% of EEC samples had nondetectable
levels of MDA, and the random assignment of concentrations between 0 and
4.1 nmol to these samples may have contributed random measurement error to the
estimated O3 effects. Because MDA represents less than 1% of lipid peroxides and is
present at low concentrations, its biological relevance has been questioned. However,
Romieu et al. (2008) pointed to their observations of statistically significant
6-89
-------
associations of EEC MDA levels with nasal lavage IL-8 levels to demonstrate its
relationship with pulmonary inflammation.
Uric acid and glutathione are ROS scavengers that are present in the airway ELF.
While the roles of these markers in the inflammatory cascade of asthma are not well
defined, they have been observed to be consumed in response to short-term O3
exposure as part of an antioxidant response in controlled human exposure and animal
studies (Section 5.3.3). Results from an epidemiologic study also indicate that a
similar antioxidant response may be induced by increases in ambient O3 exposure.
Sienra-Monge et al. (2004) found O3-associated decreases in nasal lavage levels of
uric acid and glutathione in children with asthma not supplemented with antioxidant
vitamins (Table 6-17). The magnitudes of decrease were similar for O3
concentrations lagged 2 or 3 days and averaged over 3 days.
Both controlled human exposure and toxicological studies have found O3-induced
increases in the cytokines IL-6 and IL-8 (Section 5.3.3, Section 6.2.3.1, and
Section 6.2.3.3), which are involved in initiating an influx of neutrophils, a hallmark
of O3-induced inflammation (Section 6.2.3.1). Epidemiologic studies conducted in
Mexico City, Mexico, had similar findings. Sienra-Monge et al. (2004) found that an
increase in 8-h max O3 was associated with increases in nasal lavage levels of IL-6
and IL-8 (placebo group), with larger effects estimated for lag 0-2 avg than for lag 2
or 3 O3 (Table 6-17). In another cohort of children with asthma, a 30-ppb increase in
lag 0 of 8-h max O3 was associated with a 1.6 pg/mL increase (95% CI: 1.4, 1.8) in
nasal lavage levels of IL-8 (Barraza-Villarreal et al., 2008). This study also reported
a small O3-associated decrease in EEC pH (Table 6-17). EEC pH, which is thought
to reflect the proton-buffering capacity of ammonium in airways, decreases upon
asthma exacerbation, and is negatively correlated with airway levels of
pro-inflammatory cytokines (Carpagnano et al.. 2005: Kostikas et al.. 2002: Hunt et
al.. 2000).
Albeit with limited investigation, controlled human exposure studies have found
O3-induced increases in eosinophils in adults with asthma (Section 6.2.3.1).
Eosinophils are believed to be the main effector cells that initiate and sustain
inflammation in asthma and allergy (Schmekel et al., 2001). Consistent with these
findings, in a cross-sectional study of adults with asthma in Atlanta, GA, a 3 0-ppb
increase in lag 2 of 8-h max O3 was associated with a 2.4% increase (95% CI: 0.62,
4.2) in blood eosinophils (Khatri et al., 2009). Potential confounding by weather was
not evaluated in models.
The prominent influences demonstrated for ROS and antioxidants in mediating the
respiratory effects of O3 provide biological plausibility for effect modification by
antioxidant capacity. Effect modification by antioxidant capacity has been described
for O3-associated lung function in controlled human exposure and epidemiologic
studies (Section 6.2.1.1 and Section 6.2.1.2). An epidemiologic study conducted in
Mexico City, Mexico, also found that vitamin C and E supplementation, which
potentially increase antioxidant capacity, attenuated O3-associated inflammation and
oxidative stress. Among children with asthma supplemented daily with vitamin C
6-90
-------
and E, the ratios of the geometric means of nasal lavage IL-6 and IL-8 per 30-ppb
increases in lag 0-2 avg of 8-h max O3 were approximately 1, reflecting no change
with increases in O3 concentration (Table 6-17) (Sienra-Monge et al., 2004).
The results did not clearly delineate interactions among ambient O3 concentrations,
endogenous antioxidants, and dietary antioxidants (Table 6-17). Ozone was
associated with increases in uric acid in the antioxidant group but decreases in the
placebo group across the O3 lags examined. Associations with glutathione were
similar in the two groups. In another cohort, 8-h max O3 concentrations > 38 ppb
enhanced the effects of diets high in antioxidant vitamins and/or omega-3 fatty acids
in protecting against O3-related increases in nasal lavage IL-8 (Romieu et al.. 2009).
Information on the main effects of O3 or effect modification by diet was not
presented.
The levels of several biological markers such as eNO, EEC pH, and MDA
consistently differ between groups with and without asthma and change during an
asthma exacerbation (Corradi et al., 2003; Hunt et al., 2000); however, the
magnitudes of change associated with these overt effects are not well defined.
Ozone-associated increases in interleukins and indicators of oxidative stress were
small: 1-2% increase for each 30-ppb increase in 8-h max O3 concentration
(Table 6-17). Ozone-associated increases in eNO were larger: 6-36% increase per
30-ppb increase in 8-h max ambient O3 concentration (Berhane et al., 2011; Delfino
etal.,2010a; Khatri et al., 2009; Barraza-Villarreal et al., 2008). Some studies in
populations with asthma found that increases in ambient O3 concentration (the same
lag) were associated with increases in pulmonary inflammation concurrently and
respiratory symptoms. For example, among adults with asthma in Atlanta, GA, an
increase in lag 2 ambient O3 concentration was associated with increases in eNO,
blood eosinophils, and a decrease in quality of life score, which incorporates indices
for symptoms and activity limitations (Khatri et al.. 2009). Also, among children
with asthma in Mexico City, Mexico, lag 0 O3 was associated with increases in eNO,
nasal lavage IL-8, and concurrently assessed cough but not wheeze (Barraza-
Villarreal et al.. 2008).
Children without Asthma
In the limited investigation, short-term increases in ambient O3 concentration
(8-h max or avg) were associated with increases in pulmonary inflammation in
children without asthma (Berhane et al., 2011; Barraza-Villarreal et al., 2008)
(Figure 6-11 [and Table 6-161 and Table 6-17). The study of children in Mexico City
found a larger O3-associated increase in eNO in the children without asthma than
with asthma (13.5% versus 6.2% increase per 30-ppb increase in lag 0 of 8-h max
O3) (Barraza-Villarreal et al., 2008). Ozone was associated with similar magnitudes
of change in IL-8 and EEC pH in children with and without asthma. A distinguishing
feature of this study was that 72% of children without asthma had allergies. A study
conducted in 13 Southern California communities also found that increases in
ambient O3 concentration (8-h avg, 10 a.m.-6 p.m.) were associated with increases in
eNO in children with respiratory allergy (Berhane et al.. 2011). Coherence for these
6-91
-------
epidemiologic findings is provided by observations of O3-induced allergic
inflammation in animal models of allergy (Section 6.2.3.3 and Section 6.2.6).
Berhane et al. (2011) found O3-associated increases in eNO in children without
asthma and children without respiratory allergy, providing evidence for effects on
pulmonary inflammation in healthy children. This study provided detailed
information on differences in association among various lags of 8-h avg (10 a.m.-
6 p.m.) O3. Ozone concentrations averaged over the several hours preceding eNO
collection were not significantly associated with eNO. Consistent with other studies
examining pulmonary inflammation and oxidative stress, Berhane et al. (2011) found
that relatively short lags of O3, i.e., 1 to 5 days, were associated with increases in
eNO. However, among several types of lag-based models, including unconstrained
lag models, polynomial distributed lag models, spline-based distributed lag models,
and cumulative lag models, a 23-day cumulative lag of O3 best fit the data. Among
the studies evaluated in this ISA, Berhane et al. (2011) was unique in evaluating and
finding larger respiratory effects for multi-week (e.g., 13-30 days) average O3
concentrations. A mechanism for the effects of O3 peaking with a 23-day cumulative
lag of exposure has not been delineated. Further, with examination of such long lag
periods, there is greater potential for residual confounding by weather.
Populations with Increased Outdoor Exposures
With limited investigation, increases in ambient O3 concentration were not
consistently associated with pulmonary inflammation in populations engaged in
outdoor activity or exercise. Common limitations of these studies were the small
numbers of subjects and lack of consideration for potential confounding factors.
A study in 16 adolescent long-distance runners near Atlanta, GA was noteworthy for
the daily collection of EEC and the likely greater extent to which ambient O3
concentrations represented ambient exposures because O3 concentrations were
measured during outdoor running at a site less than 1 mile from the exercise track
(Ferdinands et al., 2008). Increases in 1-h max O3 (lags 0 to 2) were associated with
increases in EEC pH, indicating O3-associated decreases in pulmonary inflammation.
Among 9 adult male runners in Sicily, Italy examined 3 days before and 20 hours
after 3 races in fall, winter, and summer, weekly average O3 concentrations (8-h avg,
7 a.m.-3 p.m.) were positively correlated with apoptosis of neutrophils (Spearman's
r = 0.70, p <0.005) and bronchial epithelial cell differential counts (Spearman's
r = 0.47, p <0.05) but not with neutrophil or macrophage cell counts or levels of the
pro-inflammatory cytokines TNF-a and IL-8 (Chimenti et al., 2009). Associations
with O3 concentrations measured during the races (mean 35 to 89 minutes) were not
examined. This study provides evidence for new endpoints; however, the
implications of findings are limited due to the lack of a rigorous statistical analysis.
In a cross-sectional study of children at camps in south Belgium, although lung
function was not associated with O3 measured at camps during outdoor activity, an
association was found for eNO (Nickmilder et al., 2007). Children at camps with lag
0 1-h max O3 concentrations >85.2 ppb had greater intraday increases in eNO
6-92
-------
compared with children at camps with O3 concentrations <51 ppb. A benchmark
dose analysis indicated that the threshold for an O3-associated increase of 4.3 ppb
eNO (their definition of increased pulmonary inflammation) was 68.6 ppb for
1-h max O3 and 56.3 ppb for 8-h max O3. While these results provide additional
evidence for O3-associated increases in pulmonary inflammation in healthy children,
they should be interpreted with caution since they were not adjusted for any potential
confounding factors and based on camp-level comparisons.
Older Adults
The panel studies examining O3-associated changes in eNO in older adults produced
contrasting findings (Figure 6-11 [and Table 6-161). The studies differed with respect
to geographic location, inclusion of healthy subjects, exposure assessment method,
and lags of O3 examined. Delfmo et al. (2010a) followed 60 older adults with
coronary artery disease in the Los Angeles, CA area for 6 weeks each during a warm
and cool season; the specific months were not specified. Ambient O3 was measured
at subjects' retirement homes, possibly reducing some exposure measurement error
due to spatial variability. Multiday averages of O3 (3- to 9-day) were associated with
increases in eNO, with effect estimates increasing with increasing number of
averaging days. In contrast with most other studies, an association was found in the
cool season but not warm season (increase in eNO per 20-ppb increase in lag 0-4 avg
of 24-h avg O3: 35.2% [95% CI: 10.9, 59.5] in cool season, -0.06% [95% CI: -14.0,
12.8] in warm season). Despite these unusual findings for the cool season, they were
similar to findings from another study of Los Angeles area adults with asthma, which
indicated an O3-associated decrease in indoor activity during the fall season
(Eiswerth et al.. 2005).
In a cool season (September-December) study conducted in older adults (ages 54-
91 years) in Steubenville, OH, Adamkiewicz et al. (2004) found that increases in O3
(1-h avg and 24-h avg before eNO collection) were associated with decreases in
eNO, reflecting decreases in pulmonary inflammation (Figure 6-11 [and
Table 6-16]). The study included healthy adults and those with asthma or COPD.
A study in a subset of these adults illustrated why it is difficult to detect effects with
central site O3 concentrations in the cool season by showing that subjects spent
> 90% of time indoors and >77% at home and had a mean 24-h avg O3 personal-
ambient ratio of 0.27 (Sarnat et al., 2006a).
Confounding in Epidemiologic Studies of Pulmonary Inflammation and
Oxidative Stress
Except where noted in the preceding text; epidemiologic studies of pulmonary
inflammation and oxidative stress accounted for potential confounding by
meteorological factors. Increases in ambient O3 concentration were associated with
pulmonary inflammation or oxidative stress in models that adjusted for temperature
and/or humidity (Delfmo et al.. 2010a: Barraza-Villarreal et al.. 2008: Romieu et al..
6-93
-------
2008). Final results from Sienra-Monge et al. (2004) and Berhane et al. (2011) were
not adjusted for temperature because associations were not altered by adjustment for
temperature. Most studies conducted over multiple seasons adjusted for season or
time trend.
In evidence limited to a small number of studies conducted in Mexico City, Mexico,
O3-associated pulmonary inflammation and oxidative stress were not found to be
confounded by PM2 5 or PM10. These studies, which analyzed 8-hour averages for
both O3 and PM, found robust associations for O3 (Barraza-Villarreal et al., 2008;
Romieu et al., 2008; Sienra-Monge et al., 2004). Ozone and PM, both measured at
central sites located within 5 km of subjects' schools or homes, were moderately
correlated (r = 0.46 - 0.54). Weak correlations have been found between personal
exposures of O3 and PM2.5 (Section 4.3.4.1). Only Romieu et al. (2008) provided
quantitative results. Lag 0 of 8-h max O3 was associated with a similar magnitude of
increase in MDA without and with adjustment for lag 0 of 8-h max PM2 5 (ratio of
geometric means for a 30-ppb increase: 1.3 [95% CI: 1.0, 1.7]). In comparison, the
O3-adjusted effect estimate for PM25 was cut in half.
Summary of Epidemiologic Studies of Pulmonary Inflammation and
Oxidative Stress
Many epidemiologic studies provided evidence that short-term increases in ambient
O3 exposure increase pulmonary inflammation and oxidative stress in children with
asthma, with evidence primarily provided by studies conducted in Mexico City.
By also finding that associations were attenuated with higher antioxidant intake,
these studies indicated that inhaled O3 may be an important source of ROS in
airways and/or may increase pulmonary inflammation via oxidative stress-mediated
mechanisms. Studies also found O3-associated increases in pulmonary inflammation
in children with allergy (Berhane et al.. 2011; Barraza-Villarreal et al.. 2008).
The limited available evidence in children and adults with increased outdoor
exposures and older adults was inconclusive. Results did not indicate confounding of
O3 associations by temperature or humidity. Copollutant models were analyzed in a
few studies conducted in Mexico City; O3 effect estimates were robust to adjustment
for moderately correlated (r = 0.46 - 0.54) PM2 5 or PMi0 (Barraza-Villarreal et al..
2008; Romieu et al.. 2008; Sienra-Monge et al.. 2004).
Ozone-associated increases in pulmonary inflammation and oxidative stress were
found in studies that used varied exposure assessment methods: measurement on site
of subjects' outdoor activity (Nickmilder et al.. 2007). average of concentrations
measured at the closest site each hour Khatri et al. (2009). measurement at a site
within 5 km of subjects' schools or homes (Barraza-Villarreal et al.. 2008; Romieu et
al.. 2008; Sienra-Monge et al.. 2004). and measurement at single site per town
(Berhane et al.. 2011). While these methods may differ in the degree of exposure
measurement error, in the limited body of evidence, there was not a clear indication
that the method of exposure assessment influenced the strength or magnitude of
associations.
6-94
-------
Most studies examined and found associations with 8-h max or daytime 8-h avg O3
concentrations, although associations also were found for 1-h max (Nickmilder et al.,
2007) and 24-h avg O3 (Delfino et al., 2010a). Collectively, studies examined
single-day O3 concentrations lagged from 0 to 5 days and concentrations averaged
over 2 to 9 days. Lag 0 of 8-h max O3 was most frequently examined and
consistently associated with pulmonary inflammation and oxidative stress. However,
in the few studies that examined multiple O3 lags, multiday average 8-h max or
8-h avg concentrations were associated with larger increases in pulmonary
inflammation and oxidative stress (Berhane et al.. 2011: Delfino et al.. 2010a: Sienra-
Monge et al.. 2004). These findings for multiday average O3 concentrations are
supported by controlled human exposure (Section 6.2.3.1) and animal studies
(Section 6.2.3.3) that similarly have found that some markers of pulmonary
inflammation remain elevated with O3 exposures repeated over multiple days.
Several epidemiologic studies concurrently examined associations of ambient O3
concentrations with biological markers of pulmonary inflammation and lung function
or respiratory symptoms. Whether evaluated at the same or different lags of O3,
associations generally were stronger for biological markers of airway inflammation
than for lung function within populations (Khatri et al., 2009; Barraza-Villarreal et
al., 2008; Nickmilder et al., 2007). Controlled human exposure studies have
demonstrated a lack of correlation between inflammatory and spirometric responses
induced by O3 exposure within subjects (Section 6.2.3.1). Evidence has suggested
that O3-related respiratory morbidity may occur via multiple mechanisms with
varying time courses of action, and the examination of a limited number of O3 lags in
these aforementioned studies may explain some of the inconsistencies in associations
of O3 with measures of pulmonary inflammation and lung function. In contrast,
based on examination in a few studies, increases in ambient O3 concentration (at the
same lag) were associated with increases in pulmonary inflammation and increases in
respiratory symptoms or activity limitations in the same population of individuals
with asthma (Khatri et al.. 2009; Barraza-Villarreal et al.. 2008).
6.2.3.3 Toxicology
The 2006 O3 AQCD states that the "extensive human clinical and animal
toxicological evidence, together with the limited available epidemiologic evidence, is
clearly indicative of a causal role for O3 in inflammatory responses in the airways"
(U.S. EPA, 2006b). Airway ciliated epithelial cells and Type 1 cells are the most O3-
sensitive cells and are initial targets of O3. These cells are damaged by O3 and
produce a number of pro-inflammatory mediators (e.g., interleukins [IL-6, IL-8],
PGE2) capable of initiating a cascade of events leading to PMN influx into the lung,
activation of alveolar macrophages, inflammation, and increased permeability across
the epithelial barrier. One critical aspect of inflammation is the potential for
metaplasia and alterations in pulmonary morphology. Studies have observed
increased thickness of the alveolar septa, presumably due to increased cellularity
after acute exposure to O3. Epithelial hyperplasia starts early in exposure and
6-95
-------
increases in magnitude for several weeks, after which it plateaus until exposure
ceases. When exposure persists for a month and longer, excess collagen and
interstitial fibrosis are observed. This response, discussed in Chapter 7, continues to
increase in magnitude throughout exposure and can even continue to increase after
exposure ends (Last et al, 1984). Previously reviewed toxicological studies of the
ability of O3 to cause inflammation, injury, and morphological changes are described
in Table 6-5 on page 6-25 (U.S. EPA. 1996f). Table 6-10 (IIS. EPA. 1996k) and
Table 6-11 (U.S. EPA. 19961) beginning on page 6-61 of the 1996 O3 AQCD, and
Annex Tables AX5-8 (U.S. EPA. 2006e) and AX5-9 (U.S. EPA. 2006f). beginning
on page AX5-17 of the 2006 O3 AQCD. Numerous recent in vitro and in vivo studies
add to this very large body of evidence for O3-induced inflammation and injury, and
provide new information regarding the underlying mechanisms (see Section 5.3).
A number of species, including dogs, rabbits, guinea pigs, rats, and mice have been
used as models to study the pulmonary effects of O3, but the similarity of non-human
primates to humans makes them an attractive model in which to study the pulmonary
response to O3. As reviewed in the 1996 and 2006 O3 AQCDs, several pulmonary
effects, including inflammation, changes in morphometry, and airway
hyperresponsiveness, have been observed in macaque and rhesus monkeys after
acute exposure to O3 (Table 6-18 presents a highlight of these studies). Increases in
inflammatory cells were observed after a single 8-hour exposure of adult rhesus
monkeys to 1 ppm O3 (Hyde et al., 1992). Inflammation was linked to morphometric
changes, such as increases in necrotic cells, smooth muscle, fibroblasts, and
nonciliated bronchiolar cells, which were observed in the trachea, bronchi, or
respiratory bronchioles. Effects have also been observed after short-term repeated
exposure to O3 at concentrations that are more relevant to ambient O3
concentrations. Morphometry changes in the lung, nose, and vocal cords were
observed after exposure to 0.15 ppm O3 for 8 hours/day for 6 days (Harkema et al..
1993: Dimitriadis. 1992: Harkema et al.. 1987a).
Since 2006, however, only one study has been published regarding acute exposure of
non-human primates to O3 (a number of recent chronic studies in non-human
primates are described in Chapter 7). In this study, a single 6-hour exposure of adult
male cynomolgus monkeys to 1 ppm O3 induced significant increases in
inflammatory and injury markers, including BAL neutrophils, total protein, alkaline
phosphatase, IL-6, IL-8, and G-CSF (Hicks et al., 2010a). Gene expression analysis
confirmed the increases in the pro-inflammatory cytokine IL-8, which had been
previously described in O3 exposed rhesus monkeys (Chang et al., 1998).
The anti-inflammatory cytokine IL-10 was also elevated, but the fold changes in
IL-10 and G-CSF were relatively low and highly variable. The single exposure also
caused necrosis and sloughing of the epithelial lining of the most distal portions of
the terminal bronchioles and the respiratory bronchioles. Bronchiolitis, alveolitis,
parenchymal and centriacinar inflammation were also observed. A second exposure
protocol (two exposures with a 2-week inter-exposure period) resulted in similar
inflammatory responses, with the exception of total protein and alkaline phosphatase
levels which were attenuated, indicating that attenuation of some but not all lavage
parameters occurred upon repeated exposure of non-human primates to O3 (Hicks et
6-96
-------
al., 2010a). This variability in attenuation is similar to the findings of earlier reports
in rodents (Wiester et al., 1996c) and non-human primates (Tyler et al., 1988).
Table 6-18 describes key morphometric studies conducted in non-human primates
exposed to O3. Morphologic observations made by Dungworth (1976) and
Dungworth et al. (1975) indicate that the rat and Bonnet monkey (Macaca radiata)
are approximately equal in susceptibility to short-term effects of O3. Mild but
discernible lesions were caused in both species by exposure to 0.2 ppm O3 for
8 hours/day for 7 days. The authors stated that detectable morphological effects in
the rat occurred at levels as low as 0.1 ppm O3. In both species, the lesion occurred at
the junction of the small airways and the gaseous exchange region. In rats, the
prominent features were accumulation of macrophages, replacement of necrotic Type
1 epithelial cells with Type 2 cells, and damage to ciliated and nonciliated Clara
cells. The principal site of damage was the alveolar duct. In monkeys, the prominent
O3-induced injury was limited to the small airways. At 0.2 ppm O3, the lesion was
observed at the proximal portion of the respiratory bronchioles. As concentrations of
O3 were increased up to 0.8 ppm, the severity of the lesion increased, and the
damage extended distally to involve the proximal portions of the alveolar duct.
Mellick et al. (1977) found similar but more pronounced effects when rhesus
monkeys (3 to 5 years of age) were exposed to 0.5 and 0.8 ppm O3, 8 hours/day for
7 days. In these experiments, the respiratory bronchioles were the most severely
damaged, whereas more distal parenchymal regions were unaffected. Major effects
were hyperplasia and hypertrophy of the nonciliated bronchiolar epithelial cells and
the accumulation of macrophages intraluminally. In mice, continuous exposure to
0.5 ppm O3 caused nodular hyperplasia of Clara cells after 7 days of exposure.
Similar findings were reported by Schwartz (1976) andSchwartz et al. (1976). who
exposed rats to 0.2, 0.5 or 0.8 ppm O3 for 8 or 24 hours/day for 1 week. Changes
observed within the proximal alveoli included infiltration of inflammatory cells and
swelling and necrosis of Type 1 cells. In the terminal bronchiole, the changes
reported were shortened cilia, clustering of basal bodies in ciliated cells suggesting
ciliogenesis, and reduction in height or loss of cytoplasmic luminal projection of the
Clara cells. Effects were seen at O3 concentrations as low as 0.2 ppm. A dose-
dependent pulmonary response to the three levels of O3 was evident. No differences
were observed in morphologic characteristics of the lesions between rats exposed
continuously and those exposed intermittently for 8 hours/day.
6-97
-------
Table 6-18 Morphometric observations in non-human primates after acute O3
exposure.
Reference
Harkema et al.
(1993)
Harkema et al.
(1987a):
Harkema et al.
(1 987b)
Dungworth
(1976)
Leonard et al.
(1991)
Chang et al.
(1998)
Hyde et al.
(1 992)
Hicks et al.
(201 Ob)
03
concentration
(ppm)
0.15
0.15
0.2
0.5
0.8
0.25
0.96
0.96
1.0
Exposure
duration
8 h/day for
6 days
8 h/day for
6 days
8 h/day for
7 days for
monkey
and rat;
continuous
at 0.5 ppm
for 7 days
for mouse
8 h/day for
7 days
8h
8h
6h
Species, Sex,
Age
Macaca radiata
(bonnet macaques)
2-6 years old
Macaca radiata,
M, F
2-6 years old
Adult Rhesus and
bonnet monkeys;
S-D rats;
Mice
Macaca radiata
age not specified
Rhesus,
M
age not specified
Rhesus,
M
2 - 8.5 years old
Cynomolgus,
M
5-7 kg
(Adult)
Observation
Several fold increase in thickness of surface
epithelium in respiratory bronchioles; increase in
interstitial mass with increase in proportion of
cuboidal cells.
Ciliated cell necrosis, shortened cilia, and
increased mucous cells in the respiratory
epithelium of nose after 0.1 5 ppm; changes in
nonciliated cells, intraepithelial leukocytes, and
mucous cells in the transitional epithelium
In both rats and monkeys mild but discernible
lesions were observed at 0.2 ppm; similar severity
between species but different site of lesions -
respiratory bronchioles for monkey and damage to
ciliated, Clara, and alveolar epithelial cells for rat;
Clara cell hyperplasia in mice
The O3 exposure level is not clear - the abstract
states 0.64 ppm, but the text mentions only
0.25 ppm. Morphometric changes in vocal cord
mucosa: disruption and hyperplasia of stratified
squamous epithelium; epithelial and connective
tissue thickness increased
Increase in IL-8 in airway epithelium correlated with
PMN influx
Increased PMNs; morphometric changes in
trachea, conducting airways, respiratory
bronchioles including increased smooth muscle in
bronchi and RB.
Increase in PMNs and IL-8 in lavage fluid
Exposure of adult BALB/c mice to 0.1 ppm O3 for 4 hours increased BAL levels of
keratinocyte chemoattractant (KC; IL-8 homologue) (~ 6-fold), IL-6 (~12-fold), and
TNF-a (~ 2-fold) (Damera et al.. 2010). Additionally, O3 increased BAL neutrophils
by 21% without changes in other cell types. A trend of increased neutrophils with
increased O3 concentration (0.12-2 ppm) was observed in BALB/c mice exposed for
3 hours (Jang et al., 2005). Although alterations in the epithelium of the airways were
not evident in 129J mice after 4 hours of exposure to 0.2 ppm O3 (Plopper et al.,
2006), detachment of the bronchiolar epithelium was observed in SD rats after 5 days
or 60 days of exposure to 0.25 ppm O3 (Oyarzun et al., 2005). Subacute (65 hours)
exposure to 0.3 ppm O3 induced pulmonary inflammation, cytokine induction, and
enhanced vascular permeability in wild type mice of a mixed background (129/Ola
and C57BL/6) and these effects were exacerbated in metallothionein I/II knockout
6-98
-------
mice (Inoue et al., 2008). Three hours or 72 hours of exposure to 0.3 ppm O3 resulted
in similar levels of IL-6 expression in the lungs of C57BL/6 mice (Johnston et al.,
2005b), along with increases in BAL protein, sTNFRl, and sTNFR2. Increased
neutrophils were observed only after the 72-hour exposure, and neither exposure
resulted in detectable levels of IL-6 or KC protein. Levels of BAL protein, sTNFRl,
and sTNFR2 were higher in the 72-hour exposure group than in the 3-hour exposure
group. In another study, the same subacute (72 hours) exposure protocol elicited
increases in BALF protein, IP-10, sTNFRl, macrophages, neutrophils, and IL-6,
IL-la, andIL-l(3 expression (Johnston et al.. 2007). Yoon et al. (2007) exposed
C57BL/6J mice continuously to 0.3 ppm O3 for 6, 24, 48, or 72 hours, and observed
elevated levels of KC, MIP-2, metalloproteinases, and inflammatory cells in the
lungs at various time points. A similar exposure protocol using C3H/HeJ and
C3H/OuJ mice demonstrated elevations in protein, PMNs, and KC, which were
predominantly TLR 4 pathway dependent based on their prominence in the TLR 4
sufficient C3H/OuJ strain Bauer et al. (2011). C3H/OuJ mice also had elevated levels
of the heat-shock protein HSP70, and further experiments in HSP70 deficient mice
indicated a role for this particular pathway in O3-related injury, discussed in more
detail in Chapter 5_.
As reviewed in the 2006 O3 AQCD, the time course for changes in BAL depends on
the parameters being studied. Similarly, after exposing adult C57BL mice to 0.5 ppm
O3 for 3 hours, Han et al. (2008) observed early (5 hours postexposure) increases in
BAL TNF-a and IL-lp, which diminished by 24 hours postexposure. Total BAL
protein was elevated at 24 hours, but there were only minimal or negligible changes
in LDH, total cells, or PMNs. Ozone increased BAL mucin levels (with statistical
significance by 24 hours postexposure), and significantly elevated surfactant protein
D at both time points. Prior intratracheal (IT) exposure to multiwalled carbon
nanotubes enhanced most of these effects, but the majority of responses to the
combined exposure were not greater than those to nanotubes alone. Ozone exposure
did not induce markers of oxidative stress in lung tissue, BAL, or serum. Consistent
with this study, Aibo et al. (2010) did not detect changes in BAL inflammatory cell
numbers in the same mouse strain after a 6-hour exposure to 0.25 or 0.5 ppm.
The majority of inflammatory cytokines (pulmonary or circulating) were not
significantly changed (as assessed 9 hours post-O3 exposure). Exposure of C57BL/6
mice to 1 ppm for 3 hours increased BAL total cells, neutrophils, and KC; these
responses were greatest at 24 hours postexposure. F2-isoprostane (8-isoprostane), a
marker of oxi dative stress, was also elevated by O3, peaking at 48 hours
postexposure (Voynow et al.. 2009).
Atopic asthma appears to be a risk factor for more severe airway inflammation
induced by experimental O3 exposure in humans (Balmes et al., 1997; Scannell et al.,
1996), and allergic animal models are often used to investigate the effects of O3 on
this potentially at-risk population. Farraj et al. (2010) exposed allergen-sensitized
adult male BALB/c mice to 0.5 ppm O3 for 5 hours once per week for 4 weeks.
Ovalbumin-sensitized mice exposed to O3 had significantly increased BAL
eosinophils by 85% and neutrophils by 103% relative to OVA sensitized mice
exposed to air, but these changes were not evident upon histopathological evaluation
6-99
-------
of the lung, and no O3 induced lesions were evident in the nasal passages. Ozone
increased BAL levels of N-acetyl-glucosaminidase (NAG; a marker of injury) and
protein. DEP co-exposure (2.0 mg/m3, nose only) inhibited these responses. These
pro-inflammatory effects in an allergic mouse model have also been observed in rats.
Wagner et al. (2007) exposed the relatively O3-resistant Brown Norway rat strain to
1 ppm O3 after sensitizing and challenging with OVA. Rats were exposed for 2 days,
and airway inflammation was assessed one day later. Filtered air for controls
contained less than 0.02 ppm O3. Histopathology indicated that O3 induced site-
specific lung lesions in the centriacinar regions, characterized by wall thickening
partly due to inflammatory cells influx. BAL neutrophils were elevated by O3 in
allergic rats, and modestly increased in non-allergic animals (not significant).
A slight (but not significant) increase in macrophages was observed, but eosinophil
numbers were not affected by O3. Soluble mediators of inflammation (Cys-LT,
MCP-1, and IL-6) were elevated by O3 in allergic animals but not non-allergic rats.
Treatment with yT, which neutralizes oxidized lipid radicals and protects lipids and
proteins from nitrosative damage, did not alter the morphologic character or severity
of the centriacinar lesions caused by O3, nor did it reduce neutrophil influx. It did,
however, significantly reduce O3-induced soluble inflammatory mediators in allergic
rats. The effects of O3 in animal models of allergic asthma are discussed in
Section 6.2.6.
In summary, a large number of toxicology studies have demonstrated that acute
exposure to O3 produces injury and inflammation in the mammalian lung, supporting
the observations in controlled human exposure studies (Section 6.2.3.1) and
epidemiologic studies (Section 6.2.3.2). These acute changes, both in inflammation
and morphology, provide a limited amount of evidence for long term sequelae of
exposure to O3. Related alterations resulting from long term exposure, such as
fibrotic changes, are discussed in Chapter 7.
6.2.4 Respiratory Symptoms and Medication Use
Controlled human exposure and toxicological studies have described modes of action
through which short-term O3 exposure may increase respiratory symptoms by
demonstrating O3-induced airway hyperresponsiveness (Section 6.2.2) and
pulmonary inflammation (Section 6.2.3.1 and Section 6.2.3.3). Epidemiologic studies
have not widely examined associations between ambient O3 concentrations and
airway hyperresponsiveness but have found O3-associated increases in pulmonary
inflammation and oxidative stress (Section 6.2.3.2). In addition to lung function
decrements, controlled human exposure studies clearly indicate O3-induced increases
in respiratory symptoms including pain on deep inspiration, shortness of breath, and
cough. This evidence is detailed in Section 6.2.1.1; however, salient observations
include an increase in respiratory symptoms with increasing concentration and
duration of O3 exposure and activity level of exposed subjects (McDonnell et al.,
1999b). Further, increases in total subjective respiratory symptoms have been
reported following 5.6 and 6.6 hours of exposure to 60 ppb O3 relative to baseline
6-100
-------
(Adams, 2006a). At 70 ppb, Schelegle et al. (2009) observed a statistically significant
Os-induced FEVi decrement of 6.1% at 6.6 hours and a significant increase in total
subjective symptoms at 5.6 and 6.6 hours. The findings for O3-induced respiratory
symptoms in controlled human exposure studies and the evidence integrated across
disciplines describing underlying modes of action provide biological plausibility for
epidemiologic associations observed between short-term increases in ambient O3
concentration and increases in respiratory symptoms.
In epidemiologic studies, respiratory symptom data typically are collected by having
subjects (or their parents) record symptoms and medication use in a diary without
direct supervision by study staff. Several limitations of symptom reports are well
recognized: recall error if not recorded daily, differences among subjects in the
interpretation of symptoms, differential reporting by subjects with and without
asthma, and occurrence in a smaller percentage of the population compared with
changes in lung function and biological markers of pulmonary inflammation.
Nonetheless, symptom diaries remain a convenient tool to collect individual-level
data from a large number of subjects and allow modeling of associations between
daily changes in O3 concentration and daily changes in respiratory morbidity over
multiple weeks or months. Importantly, most of the limitations described above are
sources of random measurement error that can bias effect estimates to the null or
increase the uncertainty around effect estimates. Furthermore, because respiratory
symptoms are associated with limitations in activity and function and are the primary
reason for using medication and seeking medical care, the evidence is directly
coherent with the associations consistently observed between increases in ambient
O3 concentration and increases in asthma ED visits (Section 6.2.7.3).
Most studies of respiratory symptoms were conducted in individuals with asthma,
and as was concluded in the 2006 O3 AQCD (U.S. EPA. 2006b. 1996a), the
collective body of epidemiologic evidence indicates that short-term increases in
ambient O3 concentration are associated with increases in respiratory symptoms in
children with asthma. Studies also found O3-associated increases in the use of asthma
medication by children. In a smaller body of studies, increases in ambient O3
concentration were associated with increases in respiratory symptoms in adults with
asthma. Ozone-associated increases in respiratory symptoms in healthy populations
were not as clearly indicated.
6.2.4.1 Children with Asthma
Respiratory Symptoms
Table 6-19 presents the locations, time periods, and ambient O3 concentrations for
studies examining respiratory symptoms and medication use in children with asthma.
The evidence supporting associations between short-term increases in ambient O3
concentration and increases in respiratory symptoms in children with asthma is
derived mostly from examination of 1-h max, 8-h max, or 8-h avg O3 concentrations
6-101
-------
and strong findings from a large body of single-region or single-city studies
(Figure 6-12 [and Table 6-201). The few available U.S. multicity studies produced
less consistent associations, but the overall body of epidemiologic evidence remains
compelling. As detailed below, because of specific methodological distinctions,
results from some multicity studies were not given greater consideration than results
from single city studies in weighing the evidence for ambient O3 exposure and
respiratory symptoms.
Similar to lung function, associations with respiratory symptoms in children with
asthma were found with ambient O3 concentrations assigned to subjects using
various methods with potentially different degrees of exposure measurement error.
As was discussed for lung function, methods included measurement of O3 on site of
and at the time of outdoor activity (Thurston et al., 1997), which is associated with
higher ambient-personal O3 correlations and ratios (Section 4.3.3); O3 concentrations
measured at sites within 5 km of subjects' home or school (Escamilla-Nufiez et al.,
2008; Romieu et al., 2006; Romieu et al., 1997; Romieu et al., 1996); O3 measured at
a single city site (Gielen et al., 1997); and O3 concentrations averaged across
multiple sites (Gent et al., 2003; Mortimer et al., 2002). In analyses with O3 averaged
across multiple sites, which were restricted to warm seasons, O3 concentrations
within the region were temporally correlated as indicated by high statewide
correlations [median r = 0.83 in Gent et al. (2003)1 or similar odds ratios for O3
averaged across all within-city monitors and that averaged from the three closest sites
(Mortimer et al., 2002). In these panel studies, the ambient concentrations averaged
across sites may have well represented the temporal variability in subjects' ambient
O3 exposures.
6-102
-------
Table 6-19 Mean and upper percentile O3 concentrations in epidemiologic
studies of respiratory symptoms, medication use, and activity
levels in children with asthma.
Study*
Thurston et
al. (1997)
Escamilla-
Nunez et al.
(2008)
Romieu et al.
(2006)
Romieu et al.
(1 997)
Romieu et al.
(1 996)
Gentetal.
(2003)
Mortimer et
al. (2002):
Mortimer et
al. (2000)
Gielen et al.
(1997)
Delfino et al.
(2003)
Rabinovitch
et al. (2004)
Schildcrout et
al. (2006)
Jalaludin et
al. (2004)
Location
CT River Valley, CT
Mexico City, Mexico
Mexico City, Mexico
Southern Mexico City,
Mexico
Northern Mexico City,
Mexico
CT, Southern MA
Bronx, East Harlem, NY;
Baltimore, MD; Washington,
DC; Detroit, Ml, Cleveland,
OH; Chicago, IL; St. Louis,
MO (NCICAS)
Amsterdam, Netherlands
Los Angeles, CA
Denver, CO
Albuquerque, NM; Baltimore,
MD; Boston, MA; Denver,
CO; San Diego, CA; Seattle,
WA; St. Louis, MO; Toronto,
ON, Canada (CAMP)
Sydney, Australia
Study Period
June 1991-
1993
July-Mar
2003-2005
Oct 1998-
Apr 2000
Apr-July 1991;
Nov1991-
Feb1992
Apr-July 1991;
Nov1991-
Feb1992
Apr-Sept 2001
June-Aug 1993
Apr-July 1995
Nov 1999-
Jan 2000
Nov-Mar
1 999-2002
May-Sept
1 994-1 995
Feb-Dec1994
03
Averaging
Time
1-h max
1-h max
8-h max
1-h max
1-h max
1-h max
8-h rolling avg
1-h max
8-h avg
(10 a.m.-
6 p.m.)
8-h max
8-h max
1-h max
1-h max
1-h max
1 5-h avg
(6 a.m. -9 p.m.)
Mean/Median
Concentration
(PPb)
83.6a
86.5
69
102
196
190
51.3, median 50.0
58.6, median 55.5
48
34.2b
17.1
25.4
28.2
Range in medians
across cities: 43.0-
65.8
12
Upper Percentile
Concentrations (ppb)
Max: 160a
NR
Max: 184
Max: 309
Max: 390
Max: 370
Max: 99.6
Max: 125.5
NR
Max: 56.5b
90th: 26.1, Max: 37
90th: 38.0, Max: 52
75th: 60, Max: 70.0
Range in 90th across
cities: 61 .5-94.7
Max: 43
6-103
-------
Study*
O'Connor et
al. (2008)
Ostro et al.
(2001 )
Mann et al.
(2010)
Just et al.
(2002)
Location
Boston, MA; Bronx,
Manhattan NY; Chicago, IL;
Dallas, TX, Seattle, WA;
Tucson, AZ (ICAS)
Los Angeles, CA
Fresno/Clovis, California
Paris, France
Study Period
Aug 1998-
July 2001
Aug-Oct 1 993
Winter-Summer
2000-2005
Apr-June 1996
03
Averaging
Time
24-h avg
1-h max
8-h max
24-h avg
Mean/Median
Concentration
(PPb)
NR
Los Angeles: 59.5
Pasadena: 95.8
49.4 (median)
30.0b
Upper Percent! le
Concentrations (ppb)
NR
Max: 130
Max: 220
75th: 69.5, Max: 120.0
Max:61.7b
*Note: Studies presented in order of first appearance in the text of this section.
NR = Not Reported, NCICAS = National Cooperative Inner-City Asthma Study, CAMP = Childhood Asthma Management Program,
ICAS = Inner City Asthma Study.
"Measured on site of subjects' outdoor activity.
bConcentrations converted from ug/m3 to ppb using the conversion factor of 0.51 assuming standard temperature (25°C) and
pressure (1 atm).
6-104
-------
Study
Symptom
O3Lag Subgroup
Aggregate of symptoms
Delfinoetal. (2003) Bothersome symptoms
Rabinovitch etal. (2004) Daytime symptoms
Schildcroutetal. (2006) Asthma symptoms
Gielenetal. (1997) LRS
URS
Mortimeret al. (2002) Morning symptoms
Mortimeret al. (2000)
0,8-h max
0,1-h max
0-2 avg
0
0-2 sum
1
2
3
4
1 -4 avg
Romieuetal. (1996)
Romieuetal. (1997)
Individual symptoms
Jalaludinetal. (2004)
O'Connor etal. (2008)
Just etal. (2002)
Ostro etal. (2001)
Escamilla-Nunezetal.
(2008)
Mann etal. (2010)
Thurston etal. (1997)
Romieuetal. (2006)
LRS
LRS
Wheeze
Wheeze/cough
Noctural cough
Wheeze
Cough
Wheeze
Chest symptoms
Difficulty breathing
0
0
0
2
1-1 9 avg
0
3
1
0
0
0-5 avg
All subjects
All subjects
All subjects
All subjects
All subjects
No medication
Cromolyn use
Beta-agonist/xanthine use
Steroid use —
Withoutallergy
With allergy
All
Fungi allergic
GSTM1 positive
GSTM1 null
GSTP1 lie/lie or Ile/Val
GSTP1 Val/Val
0.5 1 1.5 2
Odds ratio perunit increase in O3 (95% Cl)
2.5
Note: Results are presented first for aggregate indices of symptoms then for individual symptoms. Within each category, results
generally are organized in order of increasing mean ambient O3 concentration. LRS = lower respiratory symptoms, URS = upper
respiratory symptoms. Odds ratios are from single-pollutant models and are standardized to a 40-, 30-, and 20-ppb increase for
1-h max, 8-h max (or 8-h avg or 15-h avg), and 24-h avg O3 concentrations, respectively.
Figure 6-12 Associations between ambient O3 concentrations and respiratory
symptoms in children with asthma.
6-105
-------
Table 6-20 Associations between ambient O3 concentrations and respiratory
symptoms in children with asthma for studies presented in
Figure 6-12.
Study*
03
Averaging
Location/Population Time
O3 Standardized
Lag Symptom Subgroup OR (95% Cl)a
Studies examining aggregates of symptoms
Delfino et al.
(2003)
Rabinovitch
et al. (2004)
Schildcrout
et al. (2006)
Los Angeles, CA 8_h mg)<
22 children with asthma, ages 10- . ,
16 yr 1-h max
Denver, CO
86 children with asthma, ages 6- 1-h max
12yr
Albuquerque, NM; Baltimore, MD;
Boston, MA; Denver, CO; San
Diego, CA; Seattle, WA; St. Louis,
MO; Toronto, ON, Canada 1-h max
„ Bothersome
symptoms
0-2 Daytime
avg symptoms
Asthma
°~2 symptoms
Qiim
0.75 (0,
1 .09 (0,
1.32(1.
1 .08 (0.
1.01 (0.
.24, 2.30)
.39, 3.03)
,01 , 1 .74)
,89, 1.31)
,92,1.12)
990 children with asthma, ages 5-
12yr
Gielen et al.
(1997)
Mortimer et
al. (2002):
Mortimer et
al. (2000)
Amsterdam, Netherlands
61 children with asthma, ages 7- 8-h max
13yr
Bronx, East Harlem, NY; 8-h avg
Baltimore, MD; Washington, DC; (10 a m -
Detroit, Ml, Cleveland, OH; 6pm)
Chicago, IL; St. Louis, MO
846 children with asthma,
ages 4-9 yr
0
1
2
3
4
1-4
avg
LRS
URS
All subjects
All subjects
All subjects
All subjects
All subjects
sytpZs No medication use
Cromolyn use
p-agonist/xanthine use
Steroid use
Without allergy
With allergy
1 .04 (0.75, 1 .45)
1.16(1.02,1.32)
1 .06 (0.88, 1 .27)
1.21 (1.04,1.41)
1.02(0.88, 1.18)
1.19(1.02, 1.38)
1 .35 (1 .04, 1 .74)
1 .08 (0.62, 1 .87)
2.13(1.12,4.04)
1 .39 (0.98, 1 .98)
1.17(0.79,1.72)
1.59(1.00,2.52)
1 .35 (0.92, 1 .96)
Romieu et
al.(1996)
Romieu et
al. (1997)
Northern Mexico City, Mexico
71 children with asthma,
ages 5-7 yr
Southern Mexico City, Mexico
65 children with asthma,
ages 5-13 yr
1-h max
1-h max
0 LRS
0 LRS
1.07(1.02,
1.09(1.04,
1.12)
1.14)
Studies examining individual symptoms
Jalaludin et
al. (2004)
O'Connor et
al. (2008)
Sydney, Australia
125 children with asthma, mean
age 9.6yr
Boston, MA; Bronx, Manhattan
NY; Chicago, IL; Dallas, TX,
Seattle, WA; Tucson, AZ
Rfi1 rhilrlrpn with aQthma mpan
1 5-h avg
(6a.m.-
9p.m.)
24-h avg
0
Wheeze
1-19 Wheeze/
avg cough
0.93 (0.63,
1.15(0.94,
1 .02 (0.86,
1.37)
1.41)
1.21)
(SD) age 7.7 (2.0) yr
6-106
-------
Study*
Just et al.
(2002)
Ostro et al.
(2001 )
Escamilla-
Nunez et al.
(2008)
Mann et al.
(2010)
Thurston et
al.(1997)
Location/Population
Paris, France
82 children with asthma, mean
(SD)age 10.9 (2.5) yr
Los Angeles, CA
138 children with asthma, ages 6-
13yr
Mexico City, Mexico
147 children with asthma, mean
(SD)age9.6(2.1)yr
Fresno/Clovia, California
280 children with asthma, ages 6-
11 yr
CT River Valley, CT
166 children with asthma, ages 7-
13yr
03
Averaging
Time
24-h avg
1-h max
1-h max
8-h max
1-h max
03
Lag
0
3
1
0
0
Symptom
Nocturnal
cough
Wheeze
Wheeze
Wheeze
Chest
symptoms
Standardized
Subgroup OR (95% Cl)a
1.17(0.72,1.91)
0.94 (0.88, 1 .00)
1.08(1.03,1.14)
All 1.00(0.84,1.19)
Fungi allergic 1 .06 (0.84, 1 .34)
1.28(1.09, 1.51)
Romieu et
al. (2006)
Gent et al.
(2003)"
Mexico City, Mexico
151 children with asthma, mean
age 9 yr
CT, Southern MA
130 children with asthma on
maintenance medication
GSTM1 positive
0-5 Difficulty GSTM1 null
avg breathing GSTP1 lie/lie or Ile/Val
GSTP1 Val/Val
O3 <43.2 ppb
O3 43.2-51. 5 ppb
Wheeze O3 51 .6-58.8 ppb
O3 58.9-72.6 ppb
O3 > 72.7 ppb
O3 <43.2 ppb
03 43.2-51 .5 ppb
tigSs O3 51 .6-58.8 ppb
03 58.9-72.6 ppb
O3 > 72.7 ppb
1.10(0.98, 1.24)
1.17(1.02,1.33)
1 .06 (0.94, 1 .20)
1.30(1.10, 1.53)
1 .00 (reference)
1.04(0.89, 1.21)
1.16(1.00,1.35)
1.16(1.00, 1.35)
1.22(0.97,1.53)
1 .00 (reference)
1.11 (0.91,1.36)
1.01 (0.83, 1.23)
1.16(0.97,1.39)
1.31 (0.97, 1.77)
'Includes studies for Figure 6-12. plus others.
LRS = Lower respiratory symptoms, URS = Upper respiratory symptoms.
"Effect estimates are standardized to a 40, 30, and 20 ppb increase for 1-h max, 8-h max (or 8-h avg or 15-h avg), and 24-h avg O3,
respectively.
bResults not included in Figure 6-12 because results presented per quintile of ambient O3 concentration.
Among U.S. multicity studies of children with asthma, each of which examined a
different O3 averaging time, O3 was not consistently associated with increases in
respiratory symptoms (O'Connor et al., 2008; Schildcrout et al., 2006; Mortimer et
al., 2002). In the NCICAS (described in Section 6.2.1.2), which analyzed greater
than 11,000 person-days of data during one warm season, increases in most evaluated
lags of O3 (1 to 4 and 1-4 avg) were associated with increases in asthma symptoms.
A 30-ppb increase in lag 1-4 avg, of 8-h avg (10 a.m.-6 p.m.), O3 was associated
with an increase in morning asthma symptoms with an OR of 1.35 (95% CI: 1.04,
1.69) (Mortimer et al., 2002). The OR was similar in an analysis restricted to O3
concentrations <80 ppb. Associations were similarly strong for lags 2 and 4 of O3 but
weaker for lags 1 and 3 (Figure 6-12 [and Table 6-201).
6-107
-------
Like NCICAS, the U.S. multicity Childhood Asthma Management Program (CAMP,
with two cities in common with NCICAS, Table 6-19) collected daily symptom data,
analyzed data collected between May and September, and evaluated multiple lags of
O3 (Schildcrout et al., 2006). However, associations in CAMP were weaker for all
evaluated lags of O3. In meta-analyses that combined city-specific estimates, a
40-ppb increase in lag 0 of 1-h max O3 was associated with asthma symptoms with
an OR of 1.08 (95% CI: 0.89, 1.31). Odds ratios for lags 1 and 2 and the lag 0-2 sum
of O3 were between 1.0 and 1.03. In this study, data available from an average of 12
subjects per day per city were used to produce city-specific ORs. The person-days of
data contributing to each city-specific model were likely less than those of the other
multicity studies. These city-specific ORs then were combined in meta-analyses to
produce study-wide ORs. Because of these methodological details of CAMP, power
to detect associations with O3 likely was less than that for other pollutants (which
were analyzed using year-round data), other multicity studies, and several available
single-city studies.
Inconsistent associations between wheeze and nighttime asthma were reported in the
ICAS cohort (described in Section 6.2.1.2) (O'Connor et al., 2008); however, the
results are considered separately from the other available evidence because symptom
incidence was examined in association with 19-day avg (of 24-h avg) concentrations
of O3. Most evidence, whether from multi- or single-city studies, indicates
associations of respiratory symptoms with shorter lags of O3 up to a few days.
The implications of ICAS results are more limited because of a lack of a well-
characterized mode of action for respiratory symptoms resulting from longer lag
periods of O3 exposure. ICAS was precluded from examining shorter lag periods
because data were collected every 2 months on the number of days with symptoms
during the previous 2 weeks.
Several longitudinal studies conducted in different cohorts of children with asthma in
Mexico City, Mexico examined and found increases in respiratory symptoms in
association with 1-h max O3 concentrations (Escamilla-Nufiez et al., 2008; Romieu
et al.. 2006; Romieu etal.. 1997; Romieu et al.. 1996). Romieu et al. (1997); (1996)
found larger increases in symptoms in association with increases in 1-h max O3 at
lag 0 than at lag 1 or 2. Recent studies in Mexico City expanded on earlier evidence
by indicating associations with multiday averages of O3 concentrations. Romieu et
al. (2006) and Escamilla-Nufiez et al. (2008) found that ORs for associations of
ambient 1-h max O3 concentrations with respiratory symptoms and medication use
increased as the number of averaging days increased (up to lag 0-5 avg).
Studies of children with asthma examined factors that may modify symptom
responses to ambient O3 exposure but did not produce conclusive evidence. Larger
O3-associated (8-h avg [10 a.m.-6 p.m.] or 8-h max) increases in symptoms were
found in children taking asthma medication, although the specific medications
examined differed between studies. As with results for PEF, in the NCICAS
multicity cohort, O3-associated increases in morning symptoms were larger in
children taking cromolyn (used to treat asthma with allergy) or beta-
agonists/xanthines than in children taking no medication. Odds ratios were similar in
6-108
-------
children taking steroids and children taking no medication (Figure 6-12 [and
Table 6-201) (Mortimer et al., 2000). Among children with asthma in Southern
New England, O3-associated increases in symptoms were limited mostly to children
taking steroids, cromolyn, or leukotriene inhibitors for maintenance (Gent et al.,
2003).
In most studies of children with asthma, a majority of subjects (52 to 100%) had
atopy as determined by sensitization to any examined allergen. While other studies
found O3-associated increases in pulmonary inflammation in children with atopy
(Section 6.2.3.2) and in animal models of allergy (Section 6.2.3.3), evidence did not
indicate that the risk of O3-associated respiratory symptoms differed in children with
asthma with and without atopy. In the NCICAS, Mortimer et al. (2000) found that an
increase in lag 1-4 avg (8-h avg, 10 a.m.-6 p.m.) O3 was associated with a similar
increased incidence of asthma symptoms among the 79% of subjects with atopy and
the 21% of subjects without atopy (Figure 6-12 [and Table 6-201). Odds ratios for O3
did not differ by residential allergen levels. Among children with asthma in Fresno,
CA, most ORs for associations of single- and multi-day lags of 8-h max O3
concentrations (0-14 days) with wheeze were near or below 1.0 among all subjects.
Among the various O3 lags examined, increases in O3 were not consistently
associated with increases in wheeze in subjects with cat or fungi allergy either (Mann
etal..201Q).
Romieu et al. (2006) found differences in O3-associated respiratory symptoms by
genetic variants in GST enzymes, particularly, GSTP1 but less so for GSTM1.
Compared with GSTP1 He/Tie or Ile/Val subjects, larger effects were estimated for
GSTP1 Val/Val subjects (Figure 6-12 [and Table 6-201). The largest OR was found
for difficulty breathing in children with asthma who had both GSTM1 null and
GSTP1 Val/Val genotypes (OR: 1.49 [95% CI: 1.14, 1.93] per 40-ppb increase in lag
0-5 avg of 1-h max O3). These results are consistent with those described for
antioxidant capacity modifying O3-associated changes in lung function
(Section 6.2.1.2) and pulmonary inflammation [Section 6.2.3.2 for results in the same
cohort (Sienra-Monge et al., 2004)1; however, effect modification by GSTP1 variants
has not been consistent. (Romieu et al., 2006) found an O3-associated decrease in
FEVi only in children with GSTP1 He/Tie or Ile/Val genotype. Among children in
southern California, GSTP1 lie/lie was associated with greater risk of asthma onset
(Section 7.2.1). Asthma prevalence has not been consistently associated with a
particular GSTP1 genotype either (Tamer et al., 2004; Mapp et al., 2002;
Hemmingsen et al., 2001).
Asthma Medication Use
Although recent studies contributed mixed evidence, the collective body of evidence
supports associations between increases in ambient O3 concentration and increased
asthma medication use in children (Figure 6-13 [and Table 6-211). Most studies
examined and found associations with 1-h max O3 concentrations lagged 0 or 1 day;
however, associations also were found for multiday average O3 concentrations (lag
6-109
-------
0-5 avg in Romieu et al. (2006) and lags 0-2 avg and 0-4 avg in Just et al. (2002).
Within several studies, associations of O3 were similar for respiratory symptoms and
asthma medication use (Escamilla-Nufiez et al., 2008; Romieu et al., 2006;
Schildcrout et al., 2006; Jalaludin et al., 2004; Romieu et al., 1997; Thurston et al.,
1997). As an exception, Romieu et al. (1996) found that O3 was associated with an
increase in respiratory symptoms but not bronchodilator use, and Rabinovitch et al.
(2004) indicated statistically significant associations with symptoms but not
bronchodilator use (OR not reported). A few studies found higher odds of
O3-associated increases in asthma medication use than in respiratory symptoms (Just
et al.. 2002; Ostro et al.. 2001).
Study
Medication
O3Lag Subgroup
Jalaludin etal. (2004) Beta-agonist, no steroid 1
Corticosteroid
Gielen et al. (1997) Bronchodilator
Schildcrout etal. (2006) Rescue inhaler
Ostro etal. (2001) Extra medication
Thurston etal. (1997)
Romieu etal. (2006)
Romieu etal. (1996)
Romieu etal. (1997)
Beta-agonist
Bronchodilator
Bronchodilator
Bronchodilator
moderate/severe asthma
Los Angeles
0-5 avg GSTM1 positive
GSTM1 null
0-5 avg GSTP1 He/lie NeA/al
GSTP1 ValA/al
0
0
0.5 0.7 0.9 1.1 1.3 1.5
Odds ratio per unit increase in O3 (95% Cl)
Note: Results generally are presented in order of increasing mean ambient O3 concentration. Odds ratios are from single-pollutant
models and are standardized to a 40-ppb increase for 1-h max O3 and a 30-ppb increase for 8-h max or 15-h avg O3.
Figure 6-13 Associations between ambient O3 concentrations and asthma
medication use.
6-110
-------
Table 6-21 Associations between ambient O3 concentrations and asthma
medication use for studies presented in Figure 6-13.
03
Averaging
Study* Location/Population Time
Jalaludinet SVdneV. Australia 15-havg
al (2004) 125 children with asthma, (ba.m.-
mean age 9.6 yr 9p.m.)
Amsterdam, Netherlands
(19971 ~ 61 children with asthma, 8-h max
ages 7-1 Syr
Albuquerque, NM; Baltimore,
MD; Boston, MA; Denver, CO;
San Diego, CA; Seattle, WA;
S°h,'ld°rcu* St. Louis, MO; Toronto, ON, i.hmax
et al. (2006) Canada
990 children with asthma,
ages 5-1 2 yr
Los Angeles, CA
(2QQ-I ) L 138 children with asthma, 1-hmax
ages 6-13 yr
CT River Valley, CT
al (1997) ~ 166 children with asthma, 1-hmax
ages 7-1 Syr
Mexico City, Mexico
Romieu et . ,
al (2006) 151 children with asthma, i-nmax
mean age 9 yr
Northern Mexico City, Mexico
al (-1995) 71 children with asthma, 1-hmax
ages 5-7 yr
Southern Mexico City, Mexico
al ciggy) 65 children with asthma, 1 -h max
ages 5-1 Syr
Paris, France
(2QQ2)b 82 children with asthma, 24-h avg
mean (SD) age 10.9 (2.5) yr
CT, southern MA
(2003)" ' 130 children with asthma on 1-hmax
maintenance medication
'Includes studies in Fiqure 6-13, plus others.
03
Lag Medication Subgroup
Beta-agonist, no CS
1
Inhaled CS
0 Bronchodilator
0 Rescue inhaler
Moderate/severe
1 Any extra asthma
Los Angeles
0 Beta-agonist
GSTM1 positive
GSTM1 null
;v- Bronchodilator GSTR1 ||e/||e
or Ile/Val
GSTP1 Val/Val
0 Bronchodilator
0 Bronchodilator
0 Beta-agonist,
no steroid
O3 <43.2 ppb
O3 43.2-51. 5 ppb
0 Bronchodilator O3 51.6-58.8 ppb
O3 58.9-72.6 ppb
O3 > 72.7 ppb
Standardized
OR (95% Clf
1.06(0.91, 1.23)
1.06(0.97, 1.16)
1.10(0.78, 1.55)
1.01 (0.89, 1.15)
1.15(1.12, 1.19)
1.10(1.03, 1.19)
1.17(0.96, 1.44)
1.04(0.96, 1.13)
1.00(0.92, 1.09)
0.96(0.90, 1.02)
1.10(1.02, 1.19)
0.97(0.93, 1.01)
1.02(1.00, 1.05)
3.95(1.22, 12.9)
1 .00 (reference)
1.00(0.96, 1.05)
1.04(1.00, 1.09)
1.02(0.98, 1.07)
1.05(0.97, 1.13)
CS = Corticosteroid
aEffect estimates are standardized to a 40-ppb increase for 1 -h max O3, a 30-ppb increase for 8-h max or 1 5-h avg O3, and a
20-ppb increase for 24-h avg O3.
""Results not included in Fiaure 6-13. Results from Just et al. (2002) were out of ranae of other estimates, and results from Gent et
al. (2003) were presented per quintile of ambient O3 concentration.
6-111
-------
Changes in Activity
While investigation has been limited, evidence does not consistently demonstrate O3-
associated diminished activity in children with asthma (O'Connor et al., 2008;
Delfino et al., 2003). These studies examined different O3 averaging times and lags.
In the multicity ICAS cohort, O'Connor et al. (2008) found that a 20-ppb increase in
lag 1-19 avg (of 24-h avg O3) was associated with a 10% lower odds (95% CI: -26,
10) of slow play. In a small (n = 22) panel study conducted in children with asthma
in Los Angeles CA, Delfino et al. (2003) found that a 40-ppb increase in lag 0 of
1-h max O3 was associated with an increase in symptoms that interfered with daily
activity with an OR of 7.14 (95% CI: 1.18, 43.2). Several studies reported increases
in school absenteeism in children with asthma in association with increases in
ambient O3 concentration with long lag periods (14-day and 30-day distributed lags,
19-day avg) (O'Connor et al.. 2008: Gilliland et al.. 2001: Chen et al.. 2000).
Whereas Chen et al. (2000) and O'Connor et al. (2008) examined absences for any
reason, Gilliland et al. (2001) found associations with absences for respiratory
illnesses. Despite this evidence, several limitations are notable, including the lack of
a well-characterized mode of action for respiratory effects occurring with longer lag
periods of O3 exposure and the potential for residual seasonal confounding with
examination of long lag periods. In analyses of single-day lags, Gilliland et al. (2001)
found associations with O3 lagged 1 to 5 days, indicating respiratory absences may
be affected by O3 exposures with shorter lag periods.
6.2.4.2 Adults with Respiratory Disease
Within a small body of studies, several found that increases in ambient O3
concentration (8-hour or 1-h max) were associated with increases in respiratory
symptoms in adults with asthma (Khatri et al., 2009: Feo Brito et al., 2007: Ross et
al., 2002). Details from studies of respiratory symptoms in adults with respiratory
disease regarding location, time period, and ambient O3 concentrations are presented
in Table 6-22. These studies used different exposure assessment methods:
concentrations averaged from sites closest to subjects' location each hour (Khatri et
al., 2009) or concentrations measured at one (Ross et al., 2002) or multiple (Feo
Brito et al., 2007) city sites. Park et al. (2005a) found inconsistent associations for
24-h avg O3 measured at 10 city sites among the various symptoms and medication
use examined in adults with asthma in Korea during a period of dust storms. In a
study of adults with COPD in London, England, increases in lag 1 of 8-h max O3 (at
a single city site) were associated with higher odds of dyspnea and sputum changes
but lower odds of nasal discharge, wheeze, or upper respiratory symptoms (Peacock
etal..2011).
6-112
-------
Table 6-22 Mean and upper percentile O3 concentrations in epidemiologic
studies of respiratory symptoms and medication use in adults with
respiratory disease.
Study*
Khatri et al.
(2009)
Feo Brito et al.
(2007)
Eiswerth et al.
(2005)
Ross et al. (2002)
Peacock et al.
(2011)
Parket al.
(2005a)
Wiwatanadate
and Liwsrisakun
(2011)
Location
Atlanta, GA
Ciudad Real and
Puertollano, Spain
Glendora, CA
East Moline, IL
London, England
Incheon, Korea
Chiang Mai,
Thailand
Study Period
May-Sept
2003, 2005,
2006
May-June
2000-2001
Oct-Nov 1 983
Apr-Oct 1994
All-year 1995-
1997
Mar-June 2002
Aug 2005-
June 2006
03
Averaging
Time
8-h max
1-h max
1-h max
8-h avg
8-h max
24-h avg
24-h avg
Mean/Median
Concentration
(PPb)
61 (median)3
65.9 (Ciudad Real)"
56.8 (Puertollano)"
NR
41.5
15.5
Dust event days: 23.6
Control days: 25.1
17.5
Upper Percentile
Concentrations (ppb)
75th: 74a
Max: 1 01 .5b (Ciudad
Real);138.2b
(Puertollano)
NR
Max: 78.3
Autumn/Winter Max: 32
Spring/Summer Max: 74
NR
90th: 26.8, Max: 34.7
*Note: Studies presented in order of first appearance in the text of this section.
NR = Not Reported
alndividual-level estimates were derived based on time spent in the vicinity of various O3 monitors.
""Concentrations converted from ug/m3 to ppb using the conversion factor of 0.51 assuming standard temperature (25°C) and
pressure (1 atm).
Some studies that included adults with asthma examined populations with a high
prevalence of atopy. In a study of children and adults with asthma (at least 53% with
atopy), Ross et al. (2002) found that an increase in lag 1-3 avg of 8-h max O3 was
associated with an increase in symptom score and asthma medication use. Feo Brito
et al. (2007) followed 137 adults with asthma in two central Spain cities. All subjects
had pollen allergy and were examined during pollen season. In Puertollano, O3
concentrations were obtained from four city monitors, and a 40-ppb increase in lag 3
of 1-h max O3 was associated with a 14.3% increase (95% CI: 3.6, 26.0) in the
number of subjects reporting respiratory symptoms, adjusting only for time trend.
The association was imprecise in Ciudad Real (2.3% increase [95% CI: -14, 21%]
per 40-ppb increase in lag 4 of 1-h max O3), a city characterized by lower ambient
air pollution levels and a narrower range of ambient O3 concentrations as measured
at a single site established by investigators.
Cross-sectional studies reported ambient O3-associated decreases in activity in adults
with asthma; however, due to various limitations in the collective body of evidence,
firm conclusions are not warranted. Although conducted over single seasons, the
studies did not consider potential confounding by meteorological factors. In a warm
season study in Atlanta, GA (described in Section 6.2.1.2). Khatri et al. (2009) found
6-113
-------
that a 30-ppb increase in lag 2 of 8-h max O3 was associated with a 0.69-point
decrease (95% CI: -1.3, -0.11) in the Juniper quality of life score, which incorporates
indices for symptoms, mood, and activity limitations (7-point scale). In a fall study
conducted in the Los Angeles, CA area in individuals with asthma (age 16 years and
older), Eiswerth et al. (2005) found that a 40-ppb increase in 1-h max O3 was
associated with a 2.4% (95% CI: 0.83, 4) lower probability of indoor activity but
higher probability of outdoor activity. The authors acknowledged that their findings
were unexpected and may have been influenced by lack of control for potential
confounders, but they interpreted the decrease in indoor activities as rest replacing
chores. In contrast with the aforementioned studies, a panel study of individuals with
asthma (ages 13-78 years) in Thailand found that a 20-ppb increase in lag 4 of
24-h avg O3 was associated with a 26% (95% CI: 4, 43) lower odds of symptoms that
interfered with activities (Wiwatanadate and Liwsrisakun. 2011).
6.2.4.3 Populations not Restricted to Individuals with Asthma
Locations, time periods, and ambient O3 concentrations for studies of respiratory
symptoms in populations not restricted to individuals with asthma are presented in
Table 6-23. Most studies examined children, and in contrast with lung function
results (Section 6.2.1.2), short-term increases in ambient O3 concentration were not
consistently associated with increases in respiratory symptoms in children in the
general population (Figure 6-14 [and Table 6-241). Because examination was limited,
conclusions about the effects of ambient O3 exposure on respiratory symptoms in
adults are not warranted.
6-114
-------
Table 6-23 Mean and upper percentile O3 concentrations in epidemiologic
studies of respiratory symptoms in populations not restricted to
individuals with asthma.
Study*
Neas et al.
(1 995)
Linn et al.
(1996)
Hoek and
Brunekreef
(1995)
Rodriguez et
al. (2007)
Moon et al.
(2009)
Ward et al.
(2002)
Triche et al.
(2006)
Goldetal.
(1999)
Apte et al.
(2008)
Location
Uniontown, PA
Rubidoux, Upland,
Torrance, CA
Deurne and
Enkhuizen,
Netherlands
Perth, Australia
4 cities, South
Korea
Birmingham and
Sandwell, England
Southwestern VA
Mexico City,
Mexico
Multiple U.S. cities
(NR)
Study Period
June-August
1990
September-
June 1992-1994
March-July
1989
All-year, 1996-
2003
April-May 2003
January-March,
May-July 1997
June-August
1995-1996
January-
November 1991
Winter or
summer 1994-
1998
O3 Averaging
Time
12-h avg (8 a.m.-
8p.m.)
24-h avg personal
24-h avg ambient
1-h max
24-h avg
1 -h max
8-h avg (10 a.m.-
6p.m.)
24-h avg
24-h avg
8-h max
1-h max
24-h avg
Workday avg
(8 a.m. - 5 p.m.)
24-h avg
Mean/Median
Concentration
(PPb)
50
5
23
Deurne: 57
Enkhuizen: 59
28
33
NR
Winter median: 13.0
Summer median: 22.0
35.2
54.5
60.8
52.0a
34.2b
25.5b
Upper Percentile
Concentrations
(PPb)
NR
Max: 16
Max: 53
Max: 107
Max: 114
Max: 74
Max: 95
NR
Winter Max: 33
Summer Max: 41
75th: 40.6, Max: 56.6
75th: 64.1, Max: 87.6
75th: 70.0, Max: 95.0
Max: 103a
Max: 86.2b
Max: 67.3b
*Note: Studies presented in order of first appearance in the text of this section.
NR = Not Reported.
aMeasured at subject's schools.
""Concentrations converted from ug/m3 to ppb using the conversion factor of 0.51 assuming standard temperature (25°C) and
pressure (1 atm).
Children
Evidence of O3-associated increases in respiratory symptoms in children was
inconsistent, which did not appear to be attributable to the differences in exposure
assessment method among studies [e.g., O3 measured at a single site (Linn et al..
1996: Hoek and Brunekreef. 1995). O3 averaged across multiple city sites
(Rodriguez et al.. 2007). O3 measured at sites near schools (Moon et al.. 2009: Ward
et al.. 2002)]. Some studies that found weak or inconsistent associations between
ambient O3 concentrations and respiratory symptoms in children found
O3-associated decrements in lung function (Ward et al.. 2002: Linn et al.. 1996).
In their study of healthy children in Uniontown, Pennsylvania, Neas et al. (1995)
found differences in association with respiratory symptoms between two estimates of
O3 exposure using ambient O3 measurements from one central site in town. Subjects
6-115
-------
spent a mean 5.4 hours outdoors during the 12-hour period (8 a.m.-8 p.m.) over
which O3 concentrations were averaged and symptoms were reported. Evening
cough was more strongly associated with O3 concentrations weighted by time spent
outdoors (OR: 2.20 [95% CI: 1.02, 4.75] per 30-ppb increase in lag 0 of 12-h avg O3)
than with unweighted O3 concentrations (OR: 1.36 [95% CI: 0.86, 2.13]). Time spent
outdoors has been shown to influence O3 personal-ambient ratios and correlations
(Section 4.3.3): thus, the weighted O3 concentrations may have represented personal
O3 exposures better.
Study
Ward et al. (2002)
Symptom
Lag
Subgroup
Wheeze, summer 0
Cough, summer
Hoekand Brunekreef Cough
(1995) Any symptom
Moonet al. (2009)
0
Neasetal. (1995)
Tricheetal. (2006)
URS
Evening cough
Wheeze
All subjects
Jeju Island
0
Gold etal. (1999) Phlegm
0, 24-h avg
0, 8-h max
0, 1-h max
0
0123
Odds ratio per unit increase in O3 (95% CI)
Note: Results generally are presented in increasing order of mean ambient O3 concentration. URS = Upper respiratory symptoms.
Odds ratios are from single-pollutant models and are standardized to a 40-, 30-, and 20-ppb increase for 1-h max, 8-h max (or
8-h avg or 12-h avg), and 24-h avg O3 concentrations, respectively.
Figure 6-14 Associations between ambient O3 concentrations and respiratory
symptoms in children in the general population.
6-116
-------
Table 6-24 Associations between ambient O3 concentrations and respiratory
symptoms in children in the general population for studies
represented in Figure 6-14.
Study*
Ward et al.
(2002)
Hoekand
Brunekreef
(1995)
Moon et al.
(2009)
Neaset al.
(1 995)
Triche et al.
(2006)
Goldetal.
(1999)
Linn et al.
(1996)°
Location/ Population
Birmingham and Sandwell,
England
162 children, age 9 yr
Deurne and Enkhuizen,
Netherlands
300 children, ages 7-1 1 yr
4 cities, South Korea
696 children, ages <13 yr
Uniontown, PA
83 healthy children, 4th and 5th
grades
Southwestern VA
61 infants of mothers with
asthma, age <1 yr
Mexico City, Mexico
40 children, ages 8-1 1 yr
Rubidoux, Upland, Torrance, CA
269 children, 4th and 5th grades
03
Lag
0
0
0
0
0
1
0
03
Averaging
Time
24-h avg
1-h max
8-h avg
(10 a.m.-
6 p.m.)
1 2-h avg
(8 a.m.-
8 p.m.)
24-h avg
8-h max
1 -h max
24-h avg
24-h avg
Symptom
Wheeze,
summer
Cough, summer
Cough
Any symptom
URS
Evening cough
Wheeze
Phlegm
Evening
symptom score
Standardized
Subgroup OR (95% Clf
0.69 (0.51 , 0.94)
0.98(0.80, 1.21)
0.86 (0.61 , 1 .22)
0.94(0.76,1.16)
All subjects 0.96 (0.90, 1 .03)
Jeju Island 1.11 (0.95, 1.30)
2.20(1.02,
4.75)b
2.34(1.02,5.37)
1.48(0.49,4.41)
1 .73 (0.48, 6.22)
1 .02 (1 .00, 1 .04)
-0.96 (-2.2,
0.26)
'Includes studies in Figure 6-14. plus others.
URS = Upper respiratory symptoms
"Effect estimates are standardized to a 40-, 30-, and 20-ppb increase for 1-h max, 8-h max (or 8-h avg or 12-h avg), and 24-h avg
O3, respectively.
bO3 concentrations were weighted by the proportion of time spent outdoors.
°Results not presented in Figure 6-14 because outcome is a continuous variable indicating intensity of symptoms (negative indicates
improvement in symptoms).
Several other panel studies of children, in which asthma prevalence ranged from 0 to
50%, reported null or negative associations between various averaging times and lags
of ambient O3 concentration and respiratory symptoms (Moon et al., 2009;
Rodriguez et al., 2007; Ward et al., 2002; Linnetal, 1996; Hoek and Brunekreef,
1995). Among children in Mexico City, Gold et al. (1999) reported an increase in
phlegm in association with an increase in lag 1 of 24-h avg O3 concentration
measured at schools; however, investigators acknowledged being unable to
distinguish between the effects of O3 and PMi0 due to their high correlation
(r = 0.75).
Unlike other studies that examined ambient O3 concentrations from a single
monitoring site, Triche et al. (2006) found respiratory symptoms to be associated
with O3 measured at a site that for some subjects was located >100 miles away from
home (Figure 6-14 [and Table 6-241). Subjects included infants in Southwestern VA.
Odds ratios were 46-73% larger in the group who had mothers with asthma than
6-117
-------
among all infants (Triche et al., 2006). Larger ORs were found for 24-h avg than 1-h
or 8-h max O3 concentrations, particularly for wheeze but less so for difficulty
breathing. While these results suggested that children with mothers with asthma may
be at increased risk of O3-related respiratory morbidity, the authors acknowledged
that mothers with asthma may be more likely to report symptoms in their children.
Additionally, transient wheeze, which is common in infants, may not predict
respiratory morbidity later in life. In another cohort of children with parental history
of asthma that was followed to an older age (5 years), increases in ambient O3
concentration (increment of effect estimate not reported) were not associated with
increases in respiratory symptoms (Rodriguez et al.. 2007).
Adults
A cross-sectional study of 4,200 adult workers from 100 office buildings across the
U.S. found that multiple ambient O3 metrics, including the 24-hour average, the
workday average (8 a.m.-5 p.m.), and the late workday (3-6 p.m.) average, were
associated with similar magnitudes of increase in building-related symptoms (Apte et
al., 2008). Office workers likely have a low personal-ambient O3 correlation and
ratio, thus the implications of these findings compared to those of the other
respiratory symptom studies are limited.
6.2.4.4 Confounding in Epidemiologic Studies of Respiratory
Symptoms and Medication Use
Epidemiologic evidence does not indicate that confounding by meteorological factors
or copollutant exposures fully accounts for associations observed between short-term
increases in ambient O3 concentration and respiratory symptoms and medication use.
Except where specified in the text, studies found O3-associated increases in
respiratory symptoms or medication use in statistical models that adjusted for
temperature. Thurston et al. (1997) found no independent association between
temperature and respiratory symptoms among children with asthma at summer
camps. A few studies additionally included humidity in models (Triche et al.. 2006:
Ross et al.. 2002).
Several studies that examined populations with a high prevalence of atopy found
O3-associated increases in respiratory symptoms and asthma medication use with
adjustment for daily pollen counts (Just et al., 2002; Ross et al., 2002; Gielen et al.,
1997). Gielen et al. (1997) and Ross et al. (2002) examined populations with a high
prevalence of grass pollen allergy (52% and 38%, respectively). In a study conducted
over multiple seasons, Ross et al. (2002) found a similar magnitude of association
between O3 and morning symptoms and medication use with adjustment for pollen
counts. Feo Brito et al. (2007) followed adults in central Spain specifically with
asthma and pollen allergy. In one city, O3 was associated with an increase in the
number of subjects reporting symptoms. A smaller increase was estimated for pollen.
Conversely, in another city, pollen was associated with an increased reporting of
6-118
-------
respiratory symptoms, whereas O3 was not. The results suggested that O3 and pollen
may have independent effects that vary by location, depending on the mix of ambient
pollutants.
Results from copollutant models did not indicate strong confounding by copollutants
such as PM2.5, PMio, sulfate, SO2, or NO2 (Table 6-25). Notably, studies examined
different averaging times for O3 (1-h max or 8-h avg) and copollutants (3-h avg to
24-h avg) and reported a range of correlations between O3 and copollutants, which
may complicate interpretation of copollutant model results. Information on potential
copollutant confounding of asthma medication use results was limited.
The association between O3 and bronchodilator use did not change with adjustment
for PM2.s in Gent et al. (2003) but decreased in magnitude with adjustment for
12-h avg sulfate in Thurston et al. (1997). In Thurston et al. (1997) and Gent et al.
(2003), 1-h max O3 was highly correlated with 12-h avg sulfate (r = 0.74) and
24-h avg PM2 5 (r = 0.77), respectively, making it difficult to distinguish the
independent effects of O3. Studies conducted concurrently in two areas of Mexico
City, Mexico, examined 1-h max O3 and 24-h avg PMio or PM2 5 and found robust
ORs for respiratory symptoms for both O3 and PM (Romieu et al., 1997; Romieu et
al., 1996). Romieu et al. (1997) reported a moderate correlation between 1-h max O3
and 24-h avg PMio (r = 0.47). Associations between O3 and respiratory symptoms
were observed in NCICAS in copollutant models with SO2, NO2, or PMio, which
were examined with different averaging times and lags than was O3 (Mortimer et al.,
2002) (Table 6-25). Also difficult are interpretations of the O3-associated increases
in respiratory symptoms found with adjustment for two copollutants in the same
model (i.e., PM2 5 plus NO2 or PMi0_2.5) (Escamilla-Nufiez et al., 2008; Triche et al.,
2006).
6-119
-------
Table 6-25 Associations between ambient O3 concentrations and respiratory
symptoms in single- and copollutant models.
Study
Mortimer
etal.
(2002)
Gent et
al.
(2003)"
Thurston
etal.
(1997)
Romieu
etal.
(1996)
Romieu
etal.
(1997)
Location/Population
Bronx, East Harlem, NY;
Baltimore, MD;
Washington, DC; Detroit,
Ml, Cleveland, OH;
Chicago, IL; St. Louis,
MO
846 children with
asthma, ages 4-9 yr
CT, Southern MA
130 children with asthma
on maintenance
medication
CT River Valley
166 children with
asthma, ages 7-1 3 yr
Northern Mexico City,
Mexico
71 children with asthma,
ages 5-7 yr
Southern Mexico City,
Mexico
65 children with asthma,
ages 5-13 yr
Os Metrics
8-h avg
(10a.m.-
6 p.m.)
Lag 1 -4 avg
1-h max, Lag 0
<43.2 ppb
43.2-51 .5 ppb
51 .6-58.8 ppb
58.9-72.6 ppb
> 72.7 ppb
1 -h max
LagO
1 -h max
LagO
1 -h max
LagO
Symptom
Morning
symptoms
• Wheeze
Chest
symptoms
Beta-
agonist use
Lower
respiratory
symptoms
Lower
respiratory
symptoms
OR for O3 in
Single-Pollutant
Model (95% Cl)a
8 cities with SO 2 data
1 .35 (1 .04, 1 .74)
7 cities with NO2 data
1 .25 (0.94, 1 .67)
3 cities with PMio
data
1.21 (0.61,2.41)
1 .00 (reference)
1.04(0.89, 1.21)
1.16(1.00, 1.35)
1.16(1.00, 1.35)
1 .22 (0.97, 1 .53)
1.21 (1.12, 1.31)"
1.20(1.09, 1.32)b
1.07(1.02, 1.12)
1.09(1.04, 1.14)
OR for O3 in Copollutant
Model (95% Cl)a
With lag 1-2 avg, 3-h avg SO2
1.23(0.94, 1.61)
With lag 1-6 avg, 24-h avg
NO2
1.14(0.85, 1.55)
With lag 1-2 avg, 24-h avg
PM10
1.08(0.49,2.39)
with lag 0, 24-h avg PM2.5
1 .00 (reference)
1 .05 (0.90, 1 .23)
1.18(1.00, 1.38)
1.25(1.05, 1.50)
1.47(1.13, 1.90)
With lagO, 12-h avg sulfate
1.19(1.06, 1.35)b
With lagO, 12-h avg sulfate
1 .07 (0.92, 1 .24)b
With lag 0, 24-h avg PM2.5
1.06(1.02, 1.10)
With lag 0, 24-h avg PM10
1.09(1.01, 1.19)
Results generally are presented in order of increasing mean ambient O3 concentration.
aORs are standardized to a 40- and 30-ppb increase for 1-h max and 8-h avg O3, respectively.
""Temperature not included in models.
6.2.4.5 Summary of Epidemiologic Studies of Respiratory
Symptoms and Asthma Medication Use
Comprising a majority of available evidence, single-city and -region epidemiologic
studies provide consistent evidence for the effects of short-term increases in ambient
O3 exposure on increasing respiratory symptoms and asthma medication use in
children with asthma (Figure 6-12 [and Table 6-201 and Figure 6-13 [and
Table 6-211). Evidence from the few available U.S. multicity studies is less
consistent (O'Connor et al.. 2008: Schildcrout et al.. 2006: Mortimer et al.. 2002).
However, methodological differences among studies make comparisons across the
6-120
-------
multicity studies difficult. Because of fewer person-days of data (Schildcrout et al.,
2006) or examination of 19-day averages of ambient O3 concentrations (O'Connor et
al., 2008), results from recent multicity studies were not given greater consideration
than results from single-city studies in weighing the evidence for ambient O3
exposure and respiratory symptoms in children with asthma. Findings from a small
body of studies indicate O3-associated increases in respiratory symptoms in adults
with asthma. Associations between short-term increases in ambient O3 concentration
and reduced activity in children or adults with asthma are not clearly demonstrated.
While O3-associated increases in school absenteeism were found in children with
asthma, evidence for respiratory-related absences and for O3 exposure lags shorter
than 14 days is sparse. The implications of results for multi-week averages of
ambient O3 concentrations are limited because of the lack of a well-characterized
mode of action for such lags of O3 exposure and the greater potential for residual
seasonal confounding with examination of long lag periods. Short-term increases in
ambient O3 concentration were not consistently associated with increases in
respiratory symptoms in groups comprising children with and without asthma.
Increases in respiratory symptoms and asthma medication use were associated with
increases in ambient O3 concentration assigned to subjects using various methods.
Associations were found with methods likely to represent ambient exposures better,
including O3 measured on site and at the time of children's outdoor activity
(Thurston et al., 1997) and concentrations weighted by time spent outdoors (Neas et
al., 1995). However, associations also were found with methods that varied in their
representation of ambient exposures and spatial variability in ambient concentrations,
i.e., concentrations averaged among subjects' locations each hour (Khatri et al.,
2009), measured within 5 km of schools or homes (Escamilla-Nufiez et al., 2008;
Romieu et al., 2006: Romieu et al., 1997: Romieu et al., 1996), averaged across
multiple sites (Feo Brito et al., 2007: Gent et al., 2003: Mortimer et al., 2002), and
measured at a single site (Ross et al., 2002: Gielen et al., 1997).
Associations with respiratory symptoms were demonstrated most frequently for
1-h max and 8-h max or avg O3, and within-study comparisons indicated similar ORs
for 1-h max and 8-h max O3 (Delfino et al., 2003: Gent et al., 2003). Respiratory
symptoms also were associated with 12-h avg and 24-h avg O3 (Jalaludin et al.,
2004: Gold et al., 1999: Neas et al., 1995). Epidemiologic studies examined
respiratory symptoms associated with O3 concentrations lagged 0 to 5 days and those
averaged over 2 to 19 days. While O3 at lags 0 or 1 were consistently associated with
respiratory symptoms, several studies found larger ORs for multiday averages (3- to
6-day) of O3 (Escamilla-Nufiez et al., 2008: Romieu et al., 2006: Just et al., 2002:
Mortimer et al., 2002: Ross et al., 2002). Epidemiologic findings for lagged or
multiday average O3 are supported by evidence that O3 sensitizes bronchial smooth
muscle to hyperreactivity and thus acts as a primer for subsequent exposure to
antigens such as allergens (Section 5.3.5). Many studies examined populations with
asthma with a high prevalence of atopy (52-100%). In these populations,
sensitization of airways provides a biologically plausible mode of action by which
increases in respiratory symptoms result from increases in O3 exposure after a lag or
accumulated over several days. Further support is provided by findings in controlled
6-121
-------
human exposure studies that airway hyperresponsiveness (Section 6.2.2.1) and some
indicators of inflammation (Section 6.2.3.1) remained elevated following repeated
O3 exposures and by observations from epidemiologic studies that increases in
pulmonary inflammation were associated with multiday average O3 concentrations
(Section 6.2.3.2).
Epidemiologic study results did not indicate that O3-associated increases in
respiratory symptoms are confounded by temperature, pollen, or copollutants.
In limited analysis, ambient O3 was associated with respiratory symptoms with
adjustment for copollutants, primarily PM. However, identifying the independent
effects of O3 in some studies was complicated due to the high correlations observed
between O3 and PM or different lags and averaging times examined for copollutants.
Nonetheless, the robustness of associations in some studies of individuals with
asthma with and without adjustment for ambient copollutant concentrations
combined with findings from controlled human exposure studies for the direct effect
of O3 exposure provide substantial evidence for the independent effects of short-term
ambient O3 exposure on increasing respiratory symptoms.
6.2.5 Lung Host Defenses
The mammalian respiratory tract has a number of closely integrated defense
mechanisms that, when functioning normally, provide protection from the potential
health effects attributed to exposure to a wide variety of inhaled particles and
microbes. For simplicity, these interrelated defenses can be divided into two major
parts: (1) nonspecific (transport, phagocytosis, and bactericidal activity) and (2)
specific (immunologic) defense mechanisms. A variety of sensitive and reliable
methods have been used to assess the effects of O3 on these components of the lung's
defense system to provide a better understanding of the health effects associated with
the inhalation of this pollutant. The 2006 O3 AQCD stated that animal toxicological
studies provide extensive evidence that acute O3 exposures as low as 0.08 to 0.5 ppm
can cause increases in susceptibility to infectious diseases due to modulation of lung
host defenses. Table 6-6 through Table 6-9 (U.S. EPA, 1996g, h, i, j) beginning on
page 6-41 of the 1996 O3 AQCD (U.S. EPA. 1996a). and Annex Table AX5-7 (U.S.
EPA. 2006d). beginning on page AX5-8 of the 2006 O3 AQCD (U.S. EPA. 2006b).
present studies on the effects of O3 on host defense mechanisms. This section
discusses the various components of host defenses, such as the mucociliary escalator,
the phagocytic, bactericidal, and regulatory role of the alveolar macrophages (AMs),
the adaptive immune system, and integrated mechanisms that are studied by
investigating the host's response to experimental pulmonary infections.
6.2.5.1 Mucociliary Clearance
The mucociliary system is one of the lung's primary defense mechanisms. It protects
the conducting airways by trapping and quickly removing material that has been
6-122
-------
deposited or is being cleared from the alveolar region by migrating alveolar
macrophages. Ciliary movement directs particles trapped on the overlying mucous
layer toward the pharynx, where the mucus is swallowed or expectorated.
The effectiveness of mucociliary clearance can be determined by measuring such
biological activities as the rate of transport of deposited particles; the frequency of
ciliary beating; structural integrity of the ciliated cells; and the size, number, and
distribution of mucus-secreting cells. Once this defense mechanism has been altered,
a buildup of both viable and nonviable inhaled substances can occur on the
epithelium which may jeopardize the health of the host, depending on the nature of
the uncleared substance. Impaired mucociliary clearance can result in an unwanted
accumulation of cellular secretions, increased infections, chronic bronchitis, and
complications associated with COPD. A number of previous studies with various
animal species have examined the effect of O3 exposure on mucociliary clearance
and reported morphological damage to the cells of the tracheobronchial tree from
acute and sub-chronic exposure to O3 0.2 ppm and higher. The cilia were either
completely absent or had become noticeably shorter or blunt. Once these animals
were placed in a clean-air environment, the structurally damaged cilia regenerated
and appeared normal (U.S. EPA, 1986). Based on such morphological observations,
related effects such as ciliostasis, increased mucus secretions, and a slowing of
mucociliary transport rates might be expected. However, no measurable changes in
ciliary beating activity have been reported due to O3 exposure alone. Essentially no
data are available on the effects of prolonged exposure to O3 on ciliary functional
activity or on mucociliary transport rates measured in the intact animal. In general,
functional studies of mucociliary transport have observed a delay in particle
clearance soon after acute exposure. Decreased clearance is more evident at higher
doses (1 ppm), and there is some evidence of attenuation of these effects (U.S. EPA.
1986). However, no recent studies have evaluated the effects of O3 on mucociliary
clearance.
6.2.5.2 Alveolobronchiolar Transport Mechanism
In addition to the transport of particles deposited on the mucous surface layer of the
conducting airways, particles deposited in the deep lung may be removed either up
the respiratory tract or through interstitial pathways to the lymphatic system.
The pivotal mechanism of alveolobronchiolar transport involves the movement of
AMs with phagocytized particles to the bottom of the mucociliary escalator. Failure
of the AMs to phagocytize and sequester the deposited particles from the vulnerable
respiratory membrane can lead to particle entry into the interstitial spaces. Once
lodged in the interstitium, particle removal is more difficult and, depending on the
toxic or infectious nature of the particle, its interstitial location may allow the particle
to set up a focus for pathologic processes. Although some studies show reduced early
(tracheobronchial) clearance after O3 exposure, late (alveolar) clearance of deposited
material is accelerated, presumably due to macrophage influx (which in itself can be
damaging due to proteases and oxidative reactions in these cells).
6-123
-------
6.2.5.3 Alveolar Macrophages
Within the gaseous exchange region of the lung, the first line of defense against
microorganisms and nonviable particles that reach the alveolar surface is the AM.
This resident phagocyte is responsible for a variety of activities, including the
detoxification and removal of inhaled particles, maintenance of pulmonary sterility
via destruction of microorganisms, and interaction with lymphocytes for
immunologic protection. Under normal conditions, AMs seek out particles deposited
on the alveolar surface and ingest them, thereby sequestering the particles from the
vulnerable respiratory membrane. To adequately fulfill their defense function, the
AMs must maintain active mobility, a high degree of phagocytic activity, and an
optimally functioning biochemical and enzyme system for bactericidal activity and
degradation of ingested material. As discussed in previous AQCDs, short periods of
O3 exposure can cause a reduction in the number of free AMs available for
pulmonary defense, and these AMs are more fragile, less phagocytic, and have
decreased lysosomal enzyme activities required for killing pathogens. For example,
in results from earlier work in rabbits, a 2-hour exposure to 0.1 ppm O3 inhibited
phagocytosis and a 3-hour exposure to 0.25 ppm decreased lysosomal enzyme
activities (Driscoll et al., 1987; Hurst et al., 1970). Similarly, AMs from rats exposed
to 0.1 ppm O3 for 1 or 3 weeks exhibited reduced hydrogen peroxide production
(Cohen et al., 2002). A controlled human exposure study reported decrements in the
ability of alveolar macrophages to phagocytize yeast following exposure of healthy
volunteers to 80 to 100 ppb O3 for 6.6-hour during moderate exercise (Devlin et al.,
1991). Although the percentage of phagocytosis-capable macrophages was
unchanged by O3 exposure, the number of yeast engulfed was reduced when
phagocytosis was complement-dependent. However, there was no difference in the
ability of macrophages to produce superoxide anion after O3 exposure. These results
are consistent with those from another controlled human exposure study in which no
changes in the level of lysosomal enzymes or superoxide anion production were
observed in macrophages lavaged from healthy human subjects exposed to 400 ppb
O3 for 2 hours with heavy intermittent exercise (Koren et al.. 1989). More recently,
Lav et al. (2007) observed no difference in phagocytic activity or oxidative burst
capacity in macrophages or monocytes from sputum or blood collected from healthy
volunteers after a 2-hour exposure to 400 ppb O3 with moderate intermittent
exercise. However, another study (Alexis et al.. 2009) found that oxidative burst and
phagocytic activity in macrophages increased in GSTM1 null subjects compared to
GSTM1 positive subjects, who had relatively unchanged macrophage function
parameters after an O3 exposure identical to that of Lay et al. (2007). Collectively,
these studies demonstrate that O3 can affect multiple steps or aspects required for
proper macrophage function, but any C-R relationship appears complex and
genotype may be a consideration. A few other recent studies have evaluated the
effects of O3 on macrophage function, but these are of questionable relevance due to
the use of in vitro exposure systems and amphibian animal models (Mikerov et al..
2008c: Dohm et al.. 2005: Klestadt et al.. 2005).
6-124
-------
6.2.5.4 Infection and Adaptive Immunity
General Effects on the Immune System
The effects of O3 on the immune system are complex and dependent on the exposure
regimen and the observation period. According to toxicological studies it appears that
the T-cell-dependent functions of the immune system are more affected than B-cell-
dependent functions (U.S. EPA, 2006b). Generally, there is an early
immunosuppressive effect that subsides with continued O3 exposure, resulting in
either a return to normal responses or an enhancement of immune responses.
However, this is not always the case as Aranyi et al. (1983) showed decreased T-cell
mitogen reactions in mice after subchronic (90-day) exposure to 0.1 ppm O3. Earlier
studies report changes in cell populations in lymphatic tissues (U.S. EPA, 2006b).
A more recent study in mice demonstrated that numbers of certain T-cell subsets in
the spleen were reduced after exposure to 0.6 ppm O3 (lOh/day x 15d) (Feng et al..
2006).
The inflammatory effects of O3 involve the innate immune system, and as such, O3
can affect adaptive (or acquired) immunity via alterations in antigen presentation and
costimulation by innate immune cells such as macrophages and dendritic cells.
Several recent controlled human exposure studies demonstrate increased expression
of molecules involved in antigen presentation or costimulation. Lay et al. (2007)
collected sputum monocytes from healthy volunteers exposed to 400 ppb O3 for 2
hours with moderate intermittent exercise and detected increases in HLA-DR, used to
present antigen to T-cells, and CD86, a costimulatory marker necessary for T-cell
activation. Upregulation of HLA-DR was also observed by Alexis et al. (2009) in
sputum dendritic cells and macrophages from GSTM1 null subjects exposed to
400 ppb O3 for 2 hours with moderate intermittent exercise. On airway monocytes
from healthy volunteers 24 hours after exposure to 80 ppb O3 for 6.6 hours with
moderate intermittent exercise, HLA-DR, CD86, and CD14 (a molecule involved in
bacterial endotoxin reactivity) were increased, whereas CD80, a costimulatory
molecule of more heterogeneous function, was decreased (Alexis et al.. 2010).
Patterns of expression on macrophages were similar, except that HLA-DR was found
to be significantly decreased after O3 exposure and CD86 was not significantly
altered. An increase in IL-12p70, a macrophage and dendritic cell product that
activates T-cells, was correlated with increased numbers of dendritic cells. It should
be noted that these results are reported as comparisons to baseline as there was no
clean air control (Alexis et al.. 2010: Alexis et al.. 2009). Another controlled human
exposure study reported no increase in IL-12p70 in sputum from healthy subjects or
those with atopy or atopy and asthma following a 2-hour exposure to 400 ppb O3
with intermittent moderate exercise (Hernandez et al.. 2010). Levels of HLA-DR,
CD 14 and CD86 were not increased on macrophages collected from any of these
subjects. It is difficult to compare these results to those of Lav et al. (2007) and
Alexis et al. (2010) due to differences in O3 concentration, cell type examined, and
timing of postexposure analysis.
6-125
-------
Although no controlled human exposure studies have examined the effects of O3 on
the ability to mount antigen-specific responses, upregulation of markers associated
with innate immune activation and antigen presentation could potentially enhance
adaptive immunity and increase immunologic responses to antigens. While enhanced
adaptive immunity may bolster defenses against infection, it also may enhance
allergic responses (Section 6.2.6).
In animal models, O3 has been found to alter responses to antigenic stimulation. For
example, antibody responses to a T-cell-dependent antigen were suppressed after a
56-day exposure of mice to 0.8 ppm O3, and a 14-day exposure to 0.5 ppm O3
decreased the antiviral antibody response following influenza virus infection (Jakab
and Hmieleski, 1988); the latter impairment may lead to lowered resistance to
re-infection. The immune response is highly influenced by the temporal relationship
between O3 exposure and antigenic stimulation. When O3 exposure preceded
Listeria infection, there were no effects on delayed-type hypersensitivity or splenic
lymphoproliferative responses; however, when O3 exposure occurred during or after
Listeria infection was initiated, these immune responses were suppressed (Van
Loveren et al., 1988). In another study, a reduction in mitogen activated T-cell
proliferation was observed after exposure to 0.6 ppm O3 for 15 days that could be
ameliorated by antioxidant supplementation. Antigen-specific proliferation decreased
by 60%, indicating attenuation of the acquired immunity needed for subsequent
memory responses (Feng et al., 2006). Ozone exposure also skewed the ex-vivo
cytokine responses elicited by non-specific stimulation toward inflammation,
decreasing IL-2 and increasing IFN-y. Modest decreases in immune function
assessed in the offspring of O3-exposed dams (mice) were observed by Sharkhuu et
al. (2011). The ability to mount delayed-type hypersensitivity responses was
significantly suppressed in 42 day-old offspring when dams were exposed to 0.8 or
1.2 ppm O3, but not 0.4 ppm, from gestational day 9-18. Humoral responses to
immunization with sheep red blood cells were unaffected, as were other immune
parameters such as splenic populations of CD45+ T-cells, iNKT-cells, and levels of
IFN-y, IL-4, and IL-17 in the BALF. Generally, continuous exposure to O3 impairs
immune responses for the first several days of exposure, followed by an adaptation to
O3 that allows a return of normal immune responses. Most species show little effect
of O3 exposures prior to immunization, but show a suppression of responses to
antigen in O3 exposures post-immunization.
Microbial Infection
Bacterial infection
A relatively large body of evidence shows that O3 increases susceptibility to bacterial
infections. The majority of studies in this area were conducted before the 1996 O3
AQCD was published and many are included in Table 6-9 (U.S. EPA. 1996J) on page
6-53 of that document (U.S. EPA. 1996a). Known contributing factors are impaired
mucociliary streaming, altered chemotaxis/motility, defective phagocytosis of
bacteria, decreased production of lysosomal enzymes or superoxide radicals by
6-126
-------
alveolar macrophages, and decreased IFN-y levels. In animal models of bacterial
infection, exposure to 0.08 ppm O3 increases streptococcus-induced mortality,
regardless of whether O3 exposure precedes or follows infection (Miller et al., 1978;
Coffin and Gardner, 1972; Coffin et al., 1967). Increases in mortality are due to the
infectious agent, thereby reflecting functional impairment of host defenses. Exercise
and copollutants can enhance the effects of O3 in infectivity models. Although both
mice and rats exhibit impaired bactericidal macrophage activity after O3 exposure,
mortality due to infection is only observed in mice. Additionally, although mice and
humans share many host defense mechanisms, there is little compelling evidence
from epidemiologic studies to suggest an association between O3 exposure and
decreased resistance to bacterial infection, and the etiology of respiratory infections
is not easily identified via ICD codes (Section 6.2.7.3).
Viral infection
Only a few studies, described in previous AQCDs, have examined the effects of O3
exposure on the outcome of viral respiratory infection [see Table 6-9 on page 6-53 of
the 1996 O3 AQCD (U.S. EPA. 1996J)]. Some studies show increased mortality in
animals, while others show diminished severity and increased survival time. There is
little to no evidence from studies of animals or humans to suggest that O3 increases
the incidence of respiratory viral infection in humans. In human volunteers infected
with rhino virus prior to O3 exposure (0.3 ppm for 5 consecutive days), no effect was
observed on viral titers, IFN-y production, or blood lymphocyte proliferative
responses to viral antigen (Henderson et al., 1988). In vitro cell culture studies of
human bronchial epithelial cells indicate O3-induced exacerbation of human
rhinovirus infection (Spannhake et al., 2002), but this is of limited relevance. More
recent studies on the interactions of O3 and viral infections have not been published.
Natural killer (NK) cells, which destroy virally infected cells and tumors in the lung,
appear to be inhibited by higher concentrations of O3 and either unaffected or
stimulated at lower concentrations. Several studies show decreases in NK cell
activity following acute exposures ranging from 0.8 to 1 ppm (Gilmour and Jakab,
1991; Van Loveren et al., 1990; Burleson et al., 1989). However, Van Loveren et al.
(1990) showed that a 1-week exposure to 0.2 or 0.4 ppm O3 increased NK cell
activity, and an urban pattern of exposure (base of 0.06 ppm with peaks of 0.25 ppm)
had no effect on NK cell activity after 1, 3, 13, 52, or 78 weeks of exposure
(Selgrade et al., 1990). A more recent study demonstrated a 35% reduction in NK
cell activity after exposure of mice to 0.6 ppm O3 (lOh/day x 15d) (Feng et al.,
2006). The defective IL-2 production demonstrated in this study may impair NK cell
activation. Alternatively, NK cell surface charge may be altered by ROS, decreasing
their adherence to target cells (Nakamura and Matsunaga, 1998).
6.2.5.5 Summary of Lung Host Defenses
Taken as a whole, the data clearly indicate that an acute O3 exposure impairs the host
defense capability of animals, primarily by depressing AM function and perhaps also
6-127
-------
by decreasing mucociliary clearance of inhaled particles and microorganisms.
Coupled with limited evidence from controlled human exposure studies, this suggests
that humans exposed to O3 could be predisposed to bacterial infections in the lower
respiratory tract. The seriousness of such infections may depend on how quickly
bacteria develop virulence factors and how rapidly PMNs are mobilized to
compensate for the deficit in AM function. It remains unclear how O3 might affect
antigen presentation and the costimulation required for T-cell activation, given the
mixed results from controlled human exposure studies, but there is toxicological
evidence for suppression of T-cell-dependent functions by O3, including reductions
in antigen-specific proliferation and antibody production, indicating the potential for
impaired acquired immunity and memory responses. To date, a limited number of
epidemiologic studies have examined associations between O3 exposure and hospital
admissions or ED visits for respiratory infection, pneumonia, or influenza. Results
have been mixed, and in some cases conflicting (see Section 6.2.7.2 and
Section 6.2.7.3). With the exception of influenza, it is difficult to ascertain whether
cases of respiratory infection or pneumonia are of viral or bacterial etiology. A study
that examined the association between O3 exposure and respiratory hospital
admissions in response to an increase in influenza intensity did observe an increase
in respiratory hospital admissions (Wong et al.. 2009). but information from
toxicological studies of O3 and viral infections is ambiguous.
6.2.6 Allergic and Asthma-Related Responses
Effects resulting from combined exposures to O3 and allergens have been studied in
a variety of animal species, generally as models of experimental asthma. Pulmonary
function and airways hyperresponsiveness in animal models of asthma are discussed
in Section 6.2.1.3 and Section 6.2.2.2. Previous evidence indicates that O3 exposure
skews immune responses toward an allergic phenotype. For example, Gershwin et al.
(1981) reported that O3 (0.8 and 0.5 ppm for 4 days) exposure caused a 34-fold
increase in the number of IgE (allergic antibody)-containing cells in the lungs of
mice. In general, the number of IgE-containing cells correlated positively with levels
of anaphylactic sensitivity. In humans, allergic rhinoconjunctivitis symptoms are
associated with increases in ambient O3 concentrations (Riediker et al.. 2001).
Recent controlled human exposure studies have observed O3-induced changes
indicating allergic skewing. Airway eosinophils, which participate in allergic disease
and inflammation, were observed to increase in volunteers with atopy and mild
asthma 18 hours following a 7.6-hour exposure to 160 ppb O3 with light intermittent
exercise (Peden et al.. 1997). No increase in airway eosinophils was observed 4 hours
after exposure of healthy subj ects or those with atopic or atopy and asthma to
400 ppb O3 for 2 hours with moderate intermittent exercise (Hernandez et al.. 2010).
However, subjects with atopy did exhibit increased IL-5, a cytokine involved in
eosinophil recruitment and activation, suggesting that perhaps these two studies
observed the same effect at different time points. Epidemiologic studies describe
associations between eosinophils and short- (Section 6.2.3.2) or long-term
(Section 7.2.5) O3 exposure, as do chronic exposure studies in non-human primates.
6-128
-------
Hernandez et al. (2010) also observed increased expression of high and low affinity
IgE receptors on sputum macrophages from atopic asthmatics, which may enhance
IgE-dependent inflammation. Sputum levels of IL-4 and IL-13, both pro-allergic
cytokines that aid in the production of IgE, were unaltered in all groups. The lack of
increase in IL-4 levels in sputum reported by Hernandez et al. (2010), along with
increased IL-5, is consistent with results from Bosson et al. (2003). in which IL-5
(but not IL-4 levels) increased in bronchial epithelial biopsy specimens following
exposure of subjects with atopy and mild asthma to 200 ppb O3 for 2 hours with
moderate intermittent exercise. IL-5 was not elevated in specimens obtained from
healthy (no asthma) O3-exposed subjects. Collectively, findings from these studies
suggest that O3 can induce or enhance certain components of allergic inflammation
in individuals with atopy or atopic asthma.
Ozone enhances inflammatory and allergic responses to allergen challenge in
sensitized animals. Short-term exposure (2 days) to 1 ppm O3 exacerbated allergic
rhinitis and lower airway allergic inflammation in Brown Norway rats, a rat strain
that is comparatively less sensitive to O3 than other rats or humans (Wagner et al.,
2009; 2007). OVA-sensitized rats were intranasally challenged with OVA on days 1
and 2, and exposed to 0 or 1 ppm O3 (8 hours/day) on days 4 and 5. Analysis at day 6
indicated that O3 exposure enhanced intraepithelial mucosubstances in the nose and
airways, induced cys-LTs, MCP-1, and IL-6 production in BALF, and upregulated
expression of the pro-allergic cytokines IL-5 and IL-13. These changes were not
evident in non-allergic controls. All of these responses were blunted by gamma-
tocopherol (yT; vitamin E) therapy. yT neutralizes oxidized lipid radicals, and
protects lipids and proteins from nitrosative damage from NO-derived metabolites.
Farraj et al. (2010) exposed allergen-sensitized adult male BALB/c mice to 0.5 ppm
O3 for 5 hours once per week for 4 weeks. Ozone exposure and O3/DEP (2.0 mg/m3)
co-exposure of OVA-sensitized mice elicited significantly greater serum IgE levels
than in DEP-exposed OVA-sensitized mice (98% and 89% increases, respectively).
Ozone slightly enhanced levels of BAL IL-5, but despite increases in IgE, caused a
significant decrease in BAL IL-4 levels. IL-10, IL-13, and IFN-y levels were
unaffected. Lung resistance and elastance were unaffected in allergen sensitized mice
exposed solely to 0.5 ppm O3 once a week for 4 weeks (Farrai et al.. 2010).
However, co-exposure to O3 and diesel exhaust particles increased lung resistance.
In addition to exacerbating existing allergic responses, O3 can also act as an adjuvant
to produce sensitization in the respiratory tract. In a model of murine asthma, using
OVA free of detectable endotoxin, inclusion of 1 ppm O3 during the initial exposures
to OVA (2 hours, days 1 and 6) enhanced the inflammatory and allergic responses to
subsequent allergen challenge (Hollingsworth et al., 2010). Compared to air exposed
animals, O3-exposed mice exhibited significantly higher levels of total cells,
macrophages, eosinophils, and PMNs in BALF, and increased total serum IgE. Pro-
allergic cytokines IL-4, and IL-5 were also significantly elevated, along with
pleiotropic Th2 cytokine IL-9 (associated with bronchial hyperresponsiveness) and
pro-inflammatory IL-17, produced by activated T-cells. Based on lower
inflammatory, IgE, and cytokine responses in Toll-like receptor 4 deficient mice, the
effects of O3 seem to be dependent on TLR 4 signaling, as are a number of other
6-129
-------
biological responses to O3 according to studies by Hollingsworth et al. (2004),
Kleeberger et al. (2000) and Garantziotis et al. (2010). The involvement of TLR 4,
along with its endogenous ligand, hyaluronan, in O3-induced responses described in
these studies has been corroborated by a controlled human exposure study by
Hernandez et al. (2010), who found increased TLR 4 expression and elevated levels
of hyaluronic acid in volunteers with atopy or atopic asthma exposed to 400 ppb O3.
This pathway is discussed in more detail in Chapter 5.. Examination of dendritic cells
(DCs) from the draining thoracic lymph nodes indicated that O3 did not enhance the
migration of DCs from the lungs to the lymph nodes, nor did it alter the expression of
functional DC markers such as CD40, MHC class II, or CD83. However, O3 did
increase expression of CD86, which is generally associated with Th2 responses and
was detected at higher levels on DCs from donors with allergic asthma compared to
those from healthy donors Chen et al. (2006b). Increased CD86 has also been
observed on airway cells collected from human subjects following exposure to O3 in
studies by Lav et al. (2007) and Alexis et al. (2009), but not Hernandez et al. (2010)
(study details described in Section 6.2.5.4).
Ozone exposure during gestation has modest effects on allergy and asthma related
endpoints in adult offspring. When dams were exposed to 1.2 ppm O3 (but not
0.8 ppm) from gestational day 9-18, some allergic and inflammatory responses to
OVA sensitization and challenge were reduced compared to air exposed controls.
Such responses included IgE levels and eosinophils, and were observed only in mice
that were immunized early in life (PND 3) as opposed to later (PND 42), perhaps due
to the proximity of O3 and antigen exposure. The effects of gestational O3 exposure
on immune function have not been widely studied, and although reductions in
allergic endpoints are not generally observed in association with O3, other
parameters of immune function were found to be reduced, so a more global
immunosuppression may underlie these effects.
In addition to pro-allergic effects, O3 could also make airborne allergens more
allergenic. When combined with NO2, O3 has been shown to enhance nitration of
common protein allergens, which may increase their allergenicity Franze et al.
(2005).
6.2.7 Hospital Admissions, Emergency Department Visits, and Physicians
Visits
6.2.7.1 Summary of Findings from 2006 O3 AQCD
The 2006 O3 AQCD evaluated numerous respiratory ED visits and hospital
admissions studies, which consisted primarily of time-series studies conducted in the
U.S., Canada, Europe, South America, Australia and Asia. Upon collectively
evaluating the scientific evidence, the 2006 O3 AQCD concluded that "the overall
evidence supports a causal relationship between acute ambient O3 exposures and
6-130
-------
increased respiratory morbidity resulting in increased ED visits [and hospital
admissions] during the warm season" (U.S. EPA, 2006b). This conclusion was
"strongly supported by the human clinical, animal toxicologicfal], and epidemiologic
evidence for [O3-induced] lung function decrements, increased respiratory
symptoms, airway inflammation, and airway hyperreactivity" (U.S. EPA, 2006b).
Since the completion of the 2006 O3 AQCD, relatively fewer studies conducted in
the U.S., Canada, and Europe have examined the association between short-term
exposure to ambient O3 and respiratory hospital admissions and ED visits with a
growing number of studies having been conducted in Asia. This section focuses
primarily on multicity studies because they examine the effect of O3 on respiratory-
related hospital admissions and ED visits over a large geographic area using a
consistent statistical methodology. Single-city studies that encompass a large number
of hospital admissions or ED visits, or included a long study-duration were also
evaluated because these studies have more power to detect whether an association
exists between short-term O3 exposure and respiratory hospital admissions and ED
visits compared to smaller single-city studies. Additional single-city studies were
also evaluated within this section, if they were conducted in locations not represented
by the larger single-city and multicity studies, or examined population-specific
characteristics not included in the larger studies that may modify the association
between short-term O3 exposure and respiratory-related hospital admissions or ED
visits. The remaining single-city studies identified were not evaluated in this section
due to factors such as inadequate study design or insufficient sample size.
It should be mentioned that when examining the association between short-term O3
exposure and respiratory health effects that require medical attention, it is important
to distinguish between hospital admissions and ED visits. This is because it is likely
that a small percentage of respiratory ED visits will be admitted to the hospital;
therefore, respiratory ED visits may represent potentially less serious, but more
common outcomes. As a result, in the following sections respiratory hospital
admission and ED visit studies are evaluated individually. Additionally, within each
section, results are presented as either a collection of respiratory diagnoses or as
individual diseases (e.g., asthma, COPD, pneumonia and other respiratory infections)
in order to evaluate the potential effect of short-term O3 exposure on each
respiratory-related outcome. The ICD codes (i.e., ICD-9 or ICD-10) that encompass
each of these endpoints are presented in Table 6-26 along with the air quality
characteristics of the city, or across all cities, included in each study evaluated in this
section.
6-131
-------
Table 6-26 Mean and upper percentile concentrations of respiratory-related
hospital admission and emergency department (ED) visit studies
evaluated.
Study
Katsouvanni
et al. (2009?'c
Cakmaket al.
(2006b)
Bigger! et al.
(2005)°
Dales et al.
(2006)
Lin et al.
(2008a)
Wong et al.
(2009)°
Medina-
Ramon et al.
(2006)h
Yang et al.
(2005b)
Zanobetti and
Schwartz
(2006)"
Silverman and
Ito (201 0)"
Location
90 U.S. cities
(NMMAPS)d
32 European
cities
(APHEA)d
12 Canadian
cities
10 Canadian
cities
4 Italian cities'
11 Canadian
cities
11 New York
regions
Hong Kong
36 U.S. cities
Vancouver,
Canada
Boston, MA
New York,
NY
Type of Visit
(ICD9/10)
Hospital Admissions:
NMMAPS:
All respiratory (460-519)
APHEA:
All respiratory (460-519)
12 Canadian cities:
All respiratory (460-
519)e
Hospital Admissions:
All respiratory (466,
480-486, 490, 491 , 492,
493, 494, 496)
Hospital Admissions:
All respiratory (460-519)
Hospital Admissions:
Respiratory disorders
(486, 768.9, 769, 770.8,
786, 799.0, 799.1)
Hospital Admissions:
Respiratory diseases
(466, 490-493, 496)
Hospital Admissions:
All respiratory (460-519)
COPD (490-496)
Hospital Admissions:
COPD (490-496,
excluding 493)
Pneumonia (480-487)
Hospital Admissions:
COPD (490-492, 494,
496)
Hospital Admissions:
Pneumonia (480-487)
Hospital Admissions:
Asthma (493)
Averaging
Time
1 -h max
24-h avg
8-h max
24-h avg
8-h max9
8-h max9
8-h max
24-h avg
24-h avg
8-h max
Mean
Concentration (ppbf
NMMAPS:
50th: 34.9-60.0
APHEA:
50th: 11.0-38.1
12 Canadian cities:
50th: 6.7-8.3
17.4
Warm season
(May-September):
5.7-60.0
17.0
44.1
18.8
Warm
(May-September): 45.8
Cool
(October-April): 27.6
All year: 14.1
Winter
(January-March): 13.2
Spring
(April-June): 19.4
Summer
(July-September): 13.8
Fall
(October-December):
10.0
22.4
Warm
(April-August): 41 .0
Upper Percentile
Concentrations (ppb)a
NMMAPS:
75th: 46.8-68.8
APHEA:
75th: 15.3-49.4
12 Canadian cities:
75th: 8.4-1 2.4
Max: 38.0-79.0
95th: 86. 1-90.0
Max: 107.5-1 15.1
95th: 24.9-46.0
75th: 54.0
Max: 217.0
75th: 25.9
Max: 100.3
NR
Max: 38.6
75th: 31.0
95th: 47.6
75th: 53
90th: 68
6-132
-------
Study Location
Stieb et al. 7 Canadian
(2009) cities
Tolbert et al. Atlanta, GA
(2007)
Darrow et al. Atlanta, GA
(2011 a)
Villeneuve et Alberta, CAN
al. (2007)"
Ito et al. New York,
(2007b) NY
Strickland et Atlanta, GA
al.(2010)
Mar and Seattle, WA
Koenig (2009)
Arbex et al. Sao Paulo,
(2009) Brazil
Type of Visit
(ICD9/10)
ED Visits:
Asthma (493)
COPD (490-492, 494-
496)
Respiratory infection
(464, 466, 480-487)
ED Visits:
All respiratory (460-465,
460.0,466.1,466.11,
466.19,477,480-486,
491 , 492, 493, 496,
786.07, 786.09)
ED Visits:
All respiratory (460-466,
477, 480-486, 491 , 492,
493, 496, 786.09)
ED Visits:
Asthma (493)
ED Visits:
Asthma (493)
ED Visits:
Asthma (493)
Wheeze (786.07 after
10/1/98, 786.09 before
10/1/98)
ED Visits:
Asthma (493-493.9)
ED Visits:
COPD (J40-44)
Averaging Mean
Time Concentration (ppbf
24-h avg 18.4
8-h max Warm: 53.0
8-h max Warm
(March-October):
8-h max: 53
1-h max Warm
(March-October):
1-h max: 62
24-h avg Warm
(March-October):
24-h avg: 30
Commute Warm
(March-October):
Commute: 35'
Day-time Warm
(March-October):
Day-time: 45'
Night-time Warm
(March-October):
Night-time: 14'
8-h max Summer
(April-September): 38.0
Winter
(October-March): 24.3
8-h max All year: 30.4
Warm
(April-September): 42.7
Cold
(October-March): 18.0
8-h max All year: 45.4'
Warm
(May-October): 55.2'
Cold
(November-April): 34.5'
1-h max Warm (May-October):
8-h max 1-h max: 38. 6
8-h max: 32.2
1-h max 48.8
Upper Percentile
Concentrations (ppb)a
75th: 19.3-28.6
75th: 67.0
90th: 82.1
Max: 147.5
8-h max:75th: 67
8-h max:Max: 148
1-h max:75th: 76
1-h max:Max: 180
24-h avg: 75th: 37
24-h avg:Max: 81
Commute:75th: 45
Commute:Max: 106
Day-time:75th: 58
Day-time:Max: 123
Night-time:75th: 22
Night-time:Max: 64
Summer:
75th: 46.0
Winter:
75th: 31 .5
All year:
95th: 68.0
Warm months:
95th: 77.0
Cold months:
95th: 33.0
NR
75th:
1-h max: 45.5
8-h max: 39.2
75th: 61.0
Max: 143.8
6-133
-------
Study
Orazzo et al.
(2009)°
Burra et al.
(2009)
Villeneuve et
al. (2006b)
Sinclair et al.
(2010)'
Type of Visit
Location (ICD9/10)
6 Italian cities ED Visits:
Wheezing
Toronto, Physician Visits:
Canada ED Asthma (493)
Toronto, Physician Visits:
Canada Allergic rhinitis (1 77)
Atlanta, GA Physician Visits:
Asthma
Upper respiratory
infection
Lower respiratory
infection
Averaging Mean Upper Percentile
Time Concentration (ppbf Concentrations (ppb)a
8-h maxk Summer NR
(April-September):
21.1-44.3
Winter
(October-March):
1 1 .5-27.9
1-hmax 33.3 95th: 66
Max: 121
8-h max 30.0 Max: 98.7
8-h max Total Study Period: NR
All-year: 44.0
25 mo Period:
All-year: 47.9
Warm: 61.2
Cold: 27.8
28 mo Period:
All-year: 40.7
Warm: 51.8
Cold: 26.0
"Some studies did not present an overall value for the mean, middle and/or upper percentiles of the O3 distribution; as a result, the
range of the mean, middle, and/or upper percentiles across all of the cities included in the study are presented.
bStudy only presented median concentrations.
°Study presented concentrations as ug/m3 Concentration was converted to ppb using the conversion factor of 0.51 assuming
standard temperature (25°C) and pressure (1 atm).
dA subset of the European and U.S. cities included in the mortality analyses were used in the hospital admissions analyses: 8 of the
32 European cities and 14 of 90 U.S. cities.
eHospital admission data was coded using three classifications (ICD-10-CA, ICD-9, and ICD-9-CM). Attempts were made by the
original investigators to convert diagnosis from ICD-10-CA back to ICD-9.
'Only 4 of the 8 cities included in the study collected O3 data.
9O3 measured from 10:00 a.m. to 6:00 p.m.
hOnly 35 of the 36 cities included in the analysis had O3 data.
'Commute (7:00 a.m. to 10:00 a.m., 4:00 p.m. to 7:00 p.m.); day-time (8:00 a.m. to 7:00 p.m.); Night-time (12:00 a.m. to 6:00 a.m.).
'Means represent population-weighted O3 concentrations.
kO3 measured from 8:00 a.m. to 4:00 p.m.
'This study did not report the ICD codes used for the conditions examined. The 25-month period represents August 1998-August
2000, and the 28-month period represents September 2000-December 2002. This study defined the warm months as April -
October and the cold months as November-March.
6.2.7.2 Hospital Admission Studies
Respiratory Diseases
The association between exposure to an air pollutant, such as O3, and daily
respiratory-related hospital admissions has primarily been examined using all
respiratory-related hospital admissions within the range of ICD-9 codes 460-519.
Recent studies published since the 2006 O3 AQCD (U.S. EPA. 2006b) attempt to
further examine the effect of O3 exposure on respiratory-related hospital admissions
through a multicity design that examines O3 effects across countries using a
standardized methodology; multicity studies that examine effects within one country;
6-134
-------
and multi- and single-city studies that attempt to examine potential modifiers of the
O3-respiratory-related hospital admission relationship.
The Air Pollution and Health: A European and North American Approach
(APHENA) study combined data from existing multicity study databases from
Canada, Europe (APHEA2) (Katsouvanni et al. 2001). and the U.S. (NMMAPS)
(Samet et al., 2000) in order to "develop more reliable estimates of the potential
acute effects of air pollution on human health [and] provide a common basis for [the]
comparison of risks across geographic areas" (Katsouvanni et al., 2009). In an
attempt to address both of these issues, the investigators conducted extensive
sensitivity analyses to evaluate the robustness of the results to different model
specifications (e.g., penalized splines [PS] versus natural splines [NS]) and the extent
of smoothing to control for seasonal and temporal trends. The trend analyses
consisted of subjecting the models to varying extent of smoothing selected either a
priori (i.e., 3 df/year, 8 df/year, and 12 df/year), which was selected through
exploratory analyses using between 2 and 20 df, or by using the absolute sum of the
residuals of the partial autocorrelation function (PACF). Although the investigators
did not identify the model they deemed to be the most appropriate for comparing the
results across study locations, they did specify that "overall effect estimates
(i.e., estimates pooled over several cities) tended to stabilize at high degrees of
freedom" (Katsouvanni et al., 2009). Therefore, in discussion of the results across the
three study locations below, the 8 df/year results are presented for both the PS and
NS models because: (1) 8 df/year is most consistent with the extent of temporal
adjustment used in previous and recent large multicity studies in the U.S.
(e.g., NMMAPS); (2) the risk estimates for 8 df/year and 12 df/year are comparable
for all three locations; (3) the models that used the PACF method did not report the
actual degrees of freedom chosen; and (4) the 3 df/year and the PACF method
resulted in negative O3 risk estimates, which is inconsistent with the results obtained
using more aggressive seasonal adjustments and suggests inadequate control for
seasonality. Additionally, in comparisons of results across studies in figures, only the
results from one of the spline models (i.e., NS) are presented because it has been
previously demonstrated that alternative spline models result in relatively similar
effect estimates (FJEL 2003). This observation is consistent with the results of the
APFfENA analysis that was conducted with a higher number of degrees of freedom
(e.g., > 8 df/year) to account for temporal trends.
Katsouvanni et al. (2009) examined respiratory hospital admissions for people aged
65 years and older using 1-h max O3 data. The extent of hospital admission and O3
data varied across the 3 datasets: Canadian dataset included 12 cities with data for
3 years (1993-1996) per city; European dataset included 8 cities with each city
having data for between 2 and 8 years from 1988-1997; and the U.S. dataset included
14 cities with each city having data for 4 to 10 years from 1985-1994 and 7 cities
having only summer O3 data. The investigators used a three-stage hierarchical model
to account for within-city, within region, and between region variability. Results
were presented individually for each region (Figure 6-15 [and Table 6-27]). Ozone
and PMio concentrations were weakly correlated in all locations in the summer
(ranging from r = 0.27 - 0.40), but not in the winter.
6-135
-------
In the Canadian cities, using all-year data, a 40 ppb increase in 1-h max O3
concentrations at lag 0-1 was associated with an increase in respiratory hospital
admissions of 8.9% (95% CI: 0.79, 16.8%) in aPS model and 8.1% (95% CI: 0.24,
16.8%) in aNS model (Katsouyanni et al., 2009). The results were somewhat
sensitive to the lag day selected, reduced when using a single-day lag (e.g., lag 1)
(PS: 6.0%; NS: 5.5%) and increased when using a distributed lag model (PS: 18.6%;
NS: 20.4%). When adjusted for PMi0, the magnitude of the effect estimate was
attenuated, but remained positive with it being slightly larger in the NS model (5.1%
[95% CI: -6.6, 18.6%]) compared to the PS model (3.1% [95% CI: -8.3, 15.9%]).
However, in the Canadian dataset the copollutant analysis was only conducted using
a 1-day lag. The large confidence intervals for both models could be attributed to the
reduction in days included in the copollutant analyses as a result of the every-6th-day
PM sampling schedule. When the analysis was restricted to the summer months,
stronger associations were observed between O3 and respiratory hospital admissions
across the lags examined, ranging from -22 to 37% (the study does not specify
whether these effect estimates are from a NS or PS model). Because O3
concentrations across the cities included in the Canadian dataset are low (median
concentrations ranging from 6.7-8.3 ppb [Table 6-26]). the standardized increment of
40 ppb for a 1-h max increase in O3 concentrations represents an unrealistic increase
in O3 concentrations in Canada and increases the magnitude, not direction, of the
observed risk estimate. As a result, calculating the O3 risk estimate using the 40 ppb
increment does not accurately reflect the observed risk of O3-related respiratory
hospital admissions. Although this increment adequately characterizes the
distribution of 1-h max O3 concentrations across the U.S. and European datasets, it
misrepresents the observed O3 concentrations in the Canadian dataset. As a result in
summary figures, for comparability, effect estimates from the Canadian dataset are
presented for both a 5.1 ppb increase in 1-h max O3 concentrations (i.e., an
approximate interquartile range [IQR] increase in O3 concentrations across the
Canadian cities) as well as the 40 ppb increment used throughout the ISA.
In Europe, weaker but positive associations were also observed in year round
analyses; 2.9% (95% CI: 0.63, 5.0%) in the PS model and 1.6% (95% CI: -1.7, 4.2%)
in the NS model at lag 0-1 for a 40 ppb increase in 1-h max O3 concentrations
(Katsouvanni et al.. 2009). Additionally, at lag 1, associations between O3 and
respiratory hospital admissions were also reduced, but in contrast to the lag 0-1
analysis, greater effects were observed in the NS model (2.9% [95% CI: 1.0, 4.9%])
compared to the PS model (1.5% [95% CI: -2.2, 5.4]). Unlike the Canadian analysis,
a distributed lag model provided limited evidence of an association between O3 and
respiratory hospital admissions. To compare with the Canadian results, focused on
adjustment for PMi0 at lag 1, O3 effect estimates using the European dataset were
increased in the PS model (2.5% [95% CI: 0.39-4.8%]) and remained robust in the
NS model (2.4% [95% CI: 0.08, 4.6%]). However, the European analysis also
examined the effect of adjusting for PMi0 at lag 0-1 and found results were
attenuated, but remained positive in both models (PS: 0.8% [95% CI: -2.3, 4.0%];
NS: 0.8% [95% CI: -1.8, 3.6%]). Unlike the Canadian and U.S. datasets, the
European dataset consisted of daily PM data. The investigators did not observe
stronger associations in the summer-only analyses for the European cities at lag 0-1
6-136
-------
(PS: 0.4% [95% CI: -3.2, 4.0%]; NS: 0.2% [95% CI: -3.3, 3.9%]), but did observe
some evidence for larger effects during the summer, an -2.5% increase, at lag 1 in
both models (the study does not present the extent of temporal smoothing used for
these models).
Location
U.S.
Canada
Europe
•-"&
1
1
0-1
0-1 —
DL(0-2)
0-1
1
1
J. —
la
1
la
01
- J.
O-la
ni /n 9^
UL \\J- Ł.]
DL(0-2)a
1
_L
la
01
-_L
O-la
DL(0-2)
DL(0-2)a
1
1
0-1 —
0-1 —
ni lr\ T\
i
01
• All-Year
— O —
— •
-0
•
— • Summer
— •
• All Vmr
o-
— • —
— •— All-Year
— O —
-•
•O
— • Summer
B
P
-10 -5 0 5 10 15 20 25 30 35 40
% Increase
Note: Black circles = all-year results; open circles = all-year results in copollutant model with PM10; and red circles = summer only
results. For Canada, lag days with an "a" next to them represent the risk estimates standardized to an approximate IQR of 5.1 ppb
fora 1-h max increase in O3 concentrations.
Figure 6-15 Percent increase in respiratory hospital admissions from natural
spline models with 8 df/yr for a 40 ppb increase in 1-h max O3
concentrations for each location of the APHENA study.
6-137
-------
Table 6-27 Corresponding effect estimates for Figure 6-15.
Location* Season Lag3 Copollutant
1
1 PM10
All-year 0-1
U.S. 0-1 PM10
DL(0-2)
0-1
1
1
1a
1 PM10
1a PM10
0-1
0-1 a
DL(0-2)
DL(0-2)a
1
1a
0-1
0-1 a
DL(0-2)
DL(0-2)a
1
1 PM10
All-year 0-1
Europe 0-1 PM10
DL(0-2)
1
0-1
% Increase (95% Cl)b
2.62 (0.63, 4.64)
2. 14 (-0.08, 4.40)
2.38 (0.00, 4.89)
1 .42 (-1 .33, 4.23)
3.34 (0.02-6.78)
2. 14 (-0.63, 4.97)
2.78 (-0.02, 5.71)
5.54 (-0.94, 12.4)
0.69 (-0.1 2, 1.50)a
5. 13 (-6.62, 18.6)
0.64 (-0.87, 2.20)a
8.12(0.24, 16.8)
1.00(0.03, 2.00)a
20.4 (4.07, 40.2)
2.4 (0.51, 4.40)a
21.4(15.0,29.0)
2.50(1.80, 3.30)a
32.0(18.6,47.7)
3.60(2.20, 5.1 0)a
37.1 (11.5, 67.5)
4.1 (1.40, 6.80)a
2.94(1.02,4.89)
2.38 (0.08, 4.64)
1.58 (-1.71, 4.1 5)
0.87 (-1 .79, 3.58)
0.79 (-4.46, 6.37)
2.46 (-0.63, 5.54)
0.24 (-3.32, 3.91)
*For effect estimates in Figure 6-15.
aFor Canada, lag days with an "a" next to them represent the risk estimates standardized to an approximate IQR of 5.1 ppb for a
1-h max increase in O3 concentrations.
bUnless noted, risk estimates standardized to 40 ppb for a 1-h max increase in O3 concentrations.
For the U.S. in year round analyses, the investigators reported a 1.4% (95% CI: -0.9,
3.9%) increase in the PS model and 2.4% (95% CI: 0.0, 4.9%) increase in the
NS model in respiratory hospital admissions at lag 0-1 for a 40 ppb increase in
1-h max O3 concentrations with similar results for both models at lag 1 (Katsouyanni
et al.. 2009). The distributed lag model provided results similar to those observed in
the European dataset with the PS model (1.1% [95% CI: -3.0, 5.3%]), but larger
6-138
-------
effects in the NS model (3.3% [95% CI: 0.02, 6.8%]), which is consistent with the
Canadian results. With adjustment for PMi0 using the U.S. data (i.e., every-6th-day
PM data), results were attenuated, but remained positive at lag 0-1 (PS: 0.6%
[95% CI: -2.0, 3.3%]; NS: 1.4% [95% CI: -1.3, 4.2%]) which is consistent with the
results presented for the European dataset. However, at lag 1, U.S. risk estimates
remained robust to the inclusion of PMi0 in copollutant models as was observed in
the Canadian and European datasets. Compared to the all-year analyses, the
investigators did not observe stronger associations in the summer-only analysis at
either lag 0-1 (-2.2%) or lag 1 (-2.8%) in both the PS and NS models (the study does
not present the extent of temporal smoothing used for these models).
Several additional multicity studies examined respiratory disease hospital admissions
in Canada and Europe. Cakmak et al. (2006b) evaluated the association between
ambient O3 concentrations and respiratory hospital admissions for all ages in 10
Canadian cities from April 1993 to March 2000. The primary objective of this study
was to examine the potential modification of the effect of ambient air pollution on
daily respiratory hospital admissions by education and income using a time-series
analysis conducted at the city-level. The authors calculated a pooled estimate across
cities for each pollutant using a random effects model by first selecting the lag day
with the strongest association from the city-specific models. For O3, the mean
lag day across cities that provided the strongest association and for which the pooled
effect estimate was calculated was 1.2 days. In this study, all-year O3 concentrations
were used in the analysis, and additional seasonal analyses were not conducted.
Cakmak et al. (2006b) reported a 4.4% increase (95% CI: 2.2, 6.5%) in respiratory
hospital admissions for a 20 ppb increase in 24-h avg O3 concentrations.
The investigators only examined the potential effect of confounding by other
pollutants through the use of a multipollutant model (i.e., two or more additional
pollutants included in the model), which is difficult to interpret due to the potential
multicollinearity between pollutants. Cakmak et al. (2006b) also conducted an
extensive analysis of potential modifiers, specifically sex, educational attainment,
and family income, of the association between air pollution and respiratory hospital
admissions. When stratified by sex, the increase in respiratory hospital admissions
due to short-term O3 exposure were similar in males (5.2% [95% CI: 3.0, 7.3%]) and
females (4.2% [95% CI: 1.8, 6.6%]). In addition, the examination of effect
modification by income found no consistent trend across the quartiles of family
income. However, there was evidence that individuals with an education level less
than the 9th grade were disproportionately affected by O3 exposure (4.6% [95% CI:
1.8, 7.5%]) compared to individuals that completed grades 9-13 (1.7% [95% CI: -1.9,
5.3%]), some university or trade school (1.4% [95% CI: -2.0, 5.1%]), or have a
university diploma (0.66% [95% CI: -3.3, 4.7%]). The association between O3 and
respiratory hospital admissions in individuals with an education level less than the
9th grade was the strongest association across all of the pollutants examined.
A multicity study conducted in Europe by Biggeri et al. (2005) examined the
association between short-term O3 exposure and respiratory hospital admissions for
all ages in four Italian cities from 1990 to 1999. In this study, O3 was only measured
during the warm season (May-September). The authors examined associations
6-139
-------
between daily respiratory hospital admissions and short-term O3 exposure at the city-
level using a time-series analysis. Pooled estimates were calculated by combining
city-specific estimates using fixed and random effects models. The investigators
found no evidence of an association between O3 exposure and respiratory hospital
admissions in the warm season in both the random (0.1% [95% CI: -5.2, 5.7%];
distributed lag 0-3) and fixed effects (0.1% [95% CI: -5.2, 5.7%]; distributed lag 0-3)
models for a 30 ppb increase in 8-h max O3 concentrations.
Additional studies examined associations between short-term O3 exposure and
respiratory hospital admissions specifically in children. In a multicity study
conducted in Canada, Dales et al. (2006) examined the association between all-year
ambient O3 concentrations and neonatal (ages 0-27 days) respiratory hospital
admissions in 11 Canadian cities from 1986 to 2000. The investigators used a
statistical analysis approach similar to Cakmak et al. (2006b) (i.e., time-series
analysis to examine city-specific associations, and then a random effects model to
pool estimates across cities). The authors reported that for O3, the mean lag day
across cities that provided the strongest association was 2 days. The authors reported
a 5.4% (95% CI: 2.9, 8.0%) increase in neonatal respiratory hospital admissions for a
20 ppb increase in 24-h avg O3 concentrations at lag-2 days. The results from Dales
et al. (2006) provide support for the associations observed in a smaller scale study
that examined O3 exposure and pediatric respiratory hospital admissions in
New York state (Lin et al., 2008a). Lin et al. (2008a), when examining single-day
lags of 0 to 3 days, observed a positive association between O3 and pediatric
(i.e., <18 years) respiratory admissions at lag 2 (results not presented quantitatively)
in a two-stage Bayesian hierarchical model analysis of 11 geographic regions of
New York state from 1991 to 2001. Additionally, in copollutant models with PMi0
collected every-6th day, the authors found region-specific O3 associations with
respiratory hospital admissions remained relatively robust.
Overall, the evidence from epidemiologic studies continues to support an association
between short-term O3 exposure and respiratory-related hospital admissions, but it
remains unclear whether certain factors (individual- or population-level) modify this
association. Wong et al. (2009) examined the potential modification of the
relationship between ambient O3 (along with NO2, SO2, and PM10) and respiratory
hospital admissions by influenza intensity in Hong Kong for the period 1996 - 2002.
In this study air pollution concentrations were estimated by centering non-missing
daily air pollution data on the annual mean for each monitor and then an overall daily
concentration was calculated by taking the average of the daily centered mean across
all monitors. Influenza intensity was defined as a continuous variable using the
proportion of weekly specimens positive for influenza A or B instead of defining
influenza epidemics. This approach was used to avoid any potential bias associated
with the unpredictable seasonality of influenza in Hong Kong where there are
traditionally two seasonal peaks, which is in contrast to the single peaking influenza
season in the U.S. (Wong et al., 2009). In models that examined the baseline effect
(i.e., without taking into consideration influenza intensity) of short-term O3
exposure, the authors found a 3.6% (95% CI: 1.9, 5.3%) and 3.2% (95% CI: 1.0,
5.4%) increase in respiratory hospital admissions at lag 0-1 for a 30 ppb increase in
6-140
-------
8-h max O3 concentrations for the all age and > 65 age groups, respectively. When
examining influenza intensity, Wong et al. (2009) reported that the association
between short-term exposure to O3 and respiratory hospital admissions was stronger
with higher levels of influenza intensity: additional increase in respiratory hospital
admissions above baseline of 1.4% (95% CI: 0.24, 2.6%) for all age groups and 2.4%
(95% CI: 0.94, 3.8%) for those 65 and older when influenza activity increased from
0% to 10%. No difference in effects was observed when stratifying by sex.
Cause-Specific Respiratory Outcomes
In the 2006 O3 AQCD a limited number of studies were identified that examined the
effect of short-term O3 exposure on cause-specific respiratory hospital admissions.
The limited evidence "reported positive O3 associations with... asthma and COPD,
especially... during the summer or warm season" (U.S. EPA, 2006b). Of the studies
evaluated since the completion of the 2006 O3 AQCD, more have focused on
identifying whether O3 exposure is associated with specific respiratory-related
hospital admissions, including COPD, pneumonia, and asthma, but the overall body
of evidence remains small.
Chronic Obstructive Pulmonary Disease
Medina-Ramon et al. (2006) examined the association between short-term exposure
to ambient O3 and PMi0 concentrations and Medicare hospital admissions among
individuals > 65 years of age for COPD in 35 cities in the U.S. for the years 1986-
1999. The cities included in this analysis were selected because they monitored PMi0
on a daily basis. In this study, city-specific results were obtained using a monthly
time-stratified case-crossover analysis. A meta-analysis was then conducted using
random effects models to combine the city-specific results. All cities measured O3
from May through September, while only 16 of the cities had year-round
measurements. The authors reported a 1.6% increase (95% CI: 0.48, 2.9%) in COPD
admissions for lag 0-1 in the warm season for a 30 ppb increase in 8-h max O3
concentrations. In examination of single-day lags, stronger associations were
observed for lag 1 (2.9% [95% CI: 1.8, 4.0%]) compared to lag 0 (-1.5% [95% CI:
-2.7, -0.24%]). The authors found no evidence of increased associations in cool
season (-1.9% [95% CI: -3.6, -0.06%]; lag 0-1) or year round (0.24% [95% CI: -0.78,
1.2%]; lag 0-1) analyses. In a copollutant model restricted to days in which PMi0
was available, the association between O3 and COPD hospital admissions remained
robust. Of note, the frequency of PM10 measurements varied across cities with
measurements collected either every 2, 3, or 6 days. The authors conducted
additional analyses to examine potential modification of the warm season estimates
for O3 and COPD admissions by several city-level characteristics: percentage living
in poverty, emphysema mortality rate (as an indication of smoking), daily summer
apparent temperature, and percentage of households using central air conditioning.
Of the city-level characteristics examined, stronger associations were only reported
for cities with a smaller variability in daily apparent summer temperature.
6-141
-------
In a single-city study conducted in Vancouver from 1994-1998, a location with low
ambient O3 concentrations (Table 6-26), Yang et al. (2005b) examined the
association between O3 and COPD. Ozone was moderately inversely correlated with
CO (r = -0.56), NO2 (r = -0.32), and SO2 (r = -0.34), and weakly inversely correlated
with PMio (r = -0.09), suggesting that the observed O3 effect is likely not only due to
a positive correlation with other pollutants. Yang et al. (2005b) examined 1- to 7-day
(e.g., (0-6 days) lagged moving averages and observed an 8.8% (95% CI: -12.5,
32.6%) increase in COPD admissions for lag 0-3 per 20 ppb increase in 24-h avg O3
concentrations. In two-pollutant models with every-day data for NO2, SO2, or PMi0
at lag 0-3, O3 risk estimates remained robust, but were increased slightly when CO
was added to the model (Figure 6-20 [and Table 6-291).
In the study discussed above, Wong et al. (2009) also examined the potential
modification of the relationship between ambient O3 and COPD hospital admissions
by influenza intensity. The authors also found evidence of an additional increase in
COPD admissions above baseline when influenza activity increased from 0% to 10%
of 1.0% (95% CI: -0.82, 2.9%) for all age groups and 2.4% (95% CI: 0.41, 4.4%) for
those 65 and older. The baseline increase in COPD hospital admissions at lag 0-1 for
a 30 ppb increase in 8-h max O3 concentrations was 8.5% (95% CI: 5.6, 11.4%) for
the all age and 4.2% (95% CI: 1.1, 7.3%) > 65 age groups.
Pneumonia
In addition to COPD, Medina-Ramon et al. (2006) examined the association between
short-term exposure to ambient O3 and PMi0 concentrations and Medicare hospital
admissions among individuals > 65 years of age for pneumonia (ICD-9: 480-487).
The authors reported an increase in pneumonia-hospital admissions in the warm
season (2.5% [95% CI: 1.6, 3.5%] for a 30 ppb increase in 8-h max O3
concentrations; lag 0-1). Similar to the results observed for COPD hospital
admissions, pneumonia-hospital admissions associations were stronger at lag 1 (2.6%
[95% CI: 1.8, 3.4%]) compared to lag 0 (0.06% [95% CI: -0.72, 0.78%]), and no
evidence of an association was observed in the cool season or year round. In two-
pollutant models restricted to days for which PMi0 data was available, as discussed
above, the association between O3 exposure and pneumonia-hospital admissions
remained robust (results not presented quantitatively). The authors also examined
potential effect modification of the warm season estimates for O3-related pneumonia-
hospital admissions, as was done for COPD, by several city-level characteristics.
Stronger associations were reported in cities with a lower percentage of central air
conditioning use. Across the cities examined, the percentage of households having
central air conditioning ranged from 6 to 93%. The authors found no evidence of
effect modification of the O3-pneumonia-hospital admission relationship when
examining the other city-level characteristics.
Results from a single-city study conducted in Boston, MA, did not support the results
presented by Medina-Ramon et al. (2006). Zanobetti and Schwartz (2006) examined
the association of O3 and pneumonia Medicare hospital admissions for the period
1995-1999. Ozone was weakly positively correlated with PM2 5 (r = 0.20) and
6-142
-------
weakly inversely correlated with black carbon, NO2, and CO (-0.25, -0.14, and -0.30,
respectively). In an all-year analysis, the investigators reported a 3.8% (95% CI: -7.9,
-0.1%) decrease in pneumonia admissions for a 20 ppb increase in 24-h avg O3
concentrations at lag 0 and a 6.0% (95% CI: -11.1, -1.4%) decrease for the average
of lags 0 and 1. It should be noted that the mean daily counts of pneumonia
admissions was low for this study, -14 admissions per day compared to -271
admissions per day for Medina-Ramon et al. (2006). However, in analyses with other
pollutants Zanobetti and Schwartz (2006) did observe positive associations with
pneumonia-hospital admissions, indicating that the low number of daily hospital
admission counts probably did not influence the O3 pneumonia-hospital admissions
association in this study.
Asthma
There are relatively fewer studies that examined the association between short-term
exposure to O3 and asthma hospital admissions, presumably due to the limited power
given the relative rarity of asthma hospital admissions compared to ED or physician
visits. A study from New York City examined the association of 8-h max O3
concentrations with severe acute asthma admissions (i.e., those admitted to the
Intensive Care Unit [ICU]) during the warm season in the years 1999 through 2006
(Silverman and Ito. 2010). In this study, O3 was moderately correlated with PMi0
(r = 0.59). When stratifying by age, the investigators reported positive associations
with ICU asthma admissions for the 6- to 18-year age group (26.8% [95% CI: 1.4,
58.2%] for a 30 ppb increase in 8-h max O3 concentrations at lag 0-1), but little
evidence of associations for the other age groups examined (<6 years, 19-49, 50+,
and all ages). However, positive associations were observed for each age-stratified
group and all ages for non-ICU asthma admissions, but again the strongest
association was reported for the 6- to 18-years age group (28.2% [95% CI: 15.3,
41.5%]; lag 0-1). In two-pollutant models, O3 effect estimates for both non-ICU and
ICU hospital admissions remained robust to adjustment for PM25. In an additional
analysis, using a smooth function, the authors examined whether the shape of the C-
R curve for O3 and asthma hospital admissions (i.e., both general and ICU for all
ages) is linear. To account for the potential confounding effects of PM2 5, Silverman
andlto (2010) also included a smooth function of PM25 lag 0-1. When comparing
the curve to a linear fit line the authors found that the linear fit is a reasonable
approximation of the C-R relationship between O3 and asthma hospital admissions
around and below the level of the 1997 O3 NAAQS (Figure 6-16).
6-143
-------
Ozone: All Ages
o>
o
i nun
20
40
60
Ozone
80
100
Note: The average of 0-day and 1-day lagged 8-hour O3 was used in a two-pollutant model with PM25 lag 0-1, adjusting for
temporal trends, day of the week, and immediate and delayed weather effects. The solid lines are smoothed fit data, with long
broken lines indicating 95% confidence bands. The density of lines at the bottom of the figure indicates sample size.
Source: Reprinted with permission of the American Academy of Allergy, Asthma & Immunology (Silverman and Ito. 2010).
Figure 6-16 Estimated relative risks (RRs) of asthma hospital admissions for
8-h max Os concentrations at lag 0-1 allowing for possible
nonlinear relationships using natural splines.
Averting Behavior
The studies discussed above have found consistent positive associations between
short-term O3 exposure and respiratory-related hospital admissions, however, the
strength of these associations may be underestimated due to the studies not
accounting for averting behavior. As discussed in Section 4.6.6, a recent study
(Neidell and Kinney, 2010; Neidell, 2009) conducted in Southern California
demonstrated that controlling for avoidance behavior increases O3 effect estimates
for respiratory hospital admissions, specifically for children and older adults. These
analyses show that on days where no public alert was issued warning of high O3
concentrations there was an increase in asthma hospital admissions. Although only
one study has examined averting behavior and this study is limited to the outcome of
asthma hospital admissions in one location (i.e., Los Angeles, CA) for the years
6-144
-------
1989-1997, it does provide preliminary evidence indicating that epidemiologic
studies may underestimate associations between O3 exposure and health effects by
not accounting for behavioral modification when public health alerts are issued.
6.2.7.3 Emergency Department Visit Studies
Overall, relatively fewer studies have examined the association between short-term
O3 exposure and respiratory-related ED visits, compared to hospital admissions.
In the 2006 O3 AQCD, positive, but inconsistent, associations were observed
between O3 and respiratory-related ED visits with effects generally occurring during
the warm season. Since the completion of the previous AQCD, larger studies have
been conducted, in terms of sample size, study duration, and in some cases multiple
cities, to examine the association between O3 and ED visits for all respiratory
diseases, COPD, and asthma.
Respiratory Disease
A large single-city study conducted in Atlanta, GA, by Tolbert et al. (2007). and
subsequently re-analyzed by Darrow et al. (2011 a) using different air quality data,
provides evidence for an association between short-term exposures to ambient O3
concentrations and respiratory ED visits. Tolbert et al. (2007) examined the
association between air pollution, both gaseous pollutants and PM and its
components, and respiratory disease ED visits in all ages from 1993 to 2004.
The correlations between O3 and the other pollutants examined ranged from 0.2 for
CO and SO2 to 0.5-0.6 for the PM measures. Using an a priori average of lags 0-2 for
each air pollutant examined, the authors reported a 3.9% (95% CI: 2.7, 5.2%)
increase in respiratory ED visits for a 30 ppb increase in 8-h max O3 concentrations
during the warm season [defined as March-October in Darrow et al. (201 la)1.
In copollutant models, limited to days in which data for all pollutants were available,
O3 respiratory ED visits associations with CO, NO2, and PM10, were attenuated, but
remained positive (results not presented quantitatively).
Darrow et al. (201 la) examined the same health data as Tolbert et al. (2007). but
used air quality data from one centrally located monitor instead of the average of
multiple monitors. This study primarily focused on exploring whether differences
exist in the association between O3 exposure and respiratory-related ED visits
depending on the exposure metric used (i.e., 8-h max, 1-h max, 24-h avg, commuting
period [7:00 a.m. to 10:00 a.m.; 4:00 p.m. to 7:00 p.m.], day-time [8:00 a.m. to
7:00 p.m.] and night-time [12:00 a.m. to 6:00 a.m.]). An ancillary analysis of the
spatial variability of each exposure metric conducted by Darrow et al. (201 la) found
a rather homogenous spatial distribution of O3 concentrations (r > ~0.8) as the
distance from the central monitor increased from 10 km to 60 km for all exposure
durations, except the night-time metric. The relatively high spatial correlation gives
confidence in the use of a single monitor and the resulting risk estimates. To examine
the association between the various O3 exposure metrics and respiratory ED visits,
6-145
-------
the authors conceptually used a time-stratified case-crossover framework where
control days were selected as those days within the same calendar month and
maximum temperature as the case day. However, instead of conducting a traditional
case-crossover analysis, the authors used a Poisson model with indicator variables for
each of the strata (i.e., parameters of the control days). Darrow et al. (201 la) found
using an a priori lag of 1 day, the results were somewhat variable across exposure
metrics. The strongest associations with respiratory ED visits were found when using
the 8-h max, 1-h max, and day-time exposure metrics with weaker associations using
the 24-h avg and commuting period exposure metrics; a negative association was
observed when using the night-time exposure metric (Figure 6-17). These results
indicate that using the 24-h avg exposure metric may lead to smaller O3-respiratory
ED visits risk estimates due to: (1) the dilution of relevant O3 concentrations by
averaging over hours (i.e., nighttime hours) during which O3 concentrations are
known to be low and (2) potential negative confounding by other pollutants
(e.g., CO, NO2) during the nighttime hours (Darrow et al.. 201 la).
1.03 i
Z 102-
«
a &
1.01
1 00 -
~ 0 99 -
Partial
Spearman r.
1 0.95 0.93 0.63 0.78 0.04
to
00
! 1
C3
CM
Source: Reprinted with permission of Nature Publishing Group (Darrow et al.. 2011 a).
Figure 6-17 Risk ratio for respiratory ED visits and different Os exposure
metrics in Atlanta, GA, from 1993-2004.
6-146
-------
In an additional study conducted in 6 Italian cities, Orazzo et al. (2009) examined
respiratory ED visits for ages 0-2 years in 6 Italian cities from 1996 to 2000.
However, instead of identifying respiratory ED visits using the traditional approach
of selecting ICD codes as was done by Tolbert et al. (2007) and Darrow et al.
(2011 a), Orazzo et al. (2009) used data on wheeze extracted from medical records as
an indicator of lower respiratory disease. This study examined daily counts of
wheeze in relation to air pollution using a time-stratified case-crossover approach in
which control days were matched on day of week in the same month and year as the
case day. The authors found no evidence of an association between 8-h max O3
concentrations and respiratory ED visits in children aged 0-2 years in models that
examined both single-day lags and moving averages of lags from 0-6 days in year-
round and seasonal analyses (i.e., warm and cool seasons). In all-year analyses, the
percent increase in total wheeze ranged from -1.4% to -3.3% for a 0-1 to 0-6 day lag,
respectively.
COPD
Stieb et al. (2009) also examined the association between short-term O3 exposure and
COPD ED visits in 7 Canadian cities. Across cities, in an all-year analysis, O3 was
found to be positively associated with COPD ED visits (2.4% [95% CI: -1.9, 6.9%]
at lag 1 and 4.0% [95% CI: -0.54, 8.6%] at lag 2 for a 20 ppb increase in 24-h avg O3
concentrations). In seasonal analyses, larger effects were observed between O3 and
ED visits for COPD during the warm season (i.e., April-September) 6.8% [95% CI:
0.11, 13.9%] (lag day not specified); with no associations observed in the winter
season. Stieb et al. (2009) also examined associations between respiratory-related ED
visits, including COPD, and air pollution at sub-daily time scales (i.e., 3-h avg of ED
visits versus 3-h avg pollutant concentrations) and found no evidence of consistent
associations between any pollutant and any respiratory outcome.
In a single-city study, Arbex et al. (2009) examined the association between COPD
and several ambient air pollutants, including O3, in Sao Paulo, Brazil for the years
2001-2003 for individuals over the age of 40. Associations between O3 exposure and
COPD ED visits were examined in both single-day lag (0-6 days) and polynomial
distributed lag models (0-6 days). In all-year analyses, O3 was not found to be
associated with an increase in COPD ED visits (results not presented quantitatively).
The authors also conducted stratified analyses to examine the potential modification
of the air pollutant-COPD ED visits relationship by age (e.g., 40-64, >64) and sex.
In these analyses O3 was found to have an increase in COPD ED visits for women,
but not for men or either of the age groups examined.
Asthma
In a study of 7 Canadian cities, Stieb et al. (2009) also examined the association
between exposure to air pollution (i.e., CO, NO2, O3, SO2, PM10, PM2.5, and O3) and
asthma ED visits. Associations between short-term O3 exposure and asthma ED
visits were examined at the city level and then pooled using either fixed or random
effects models depending on whether heterogeneity among effect estimates was
6-147
-------
found to be statistically significant. Across cities, in an all-year analysis, the authors
found that short-term O3 exposure was associated with an increase (4.7% [95% CI:
-1.4, 11.1%] at lag 1 and3.5% [95% CI: 0.33, 6.8%] at lag 2 for a 20 ppb increase in
24-h avg O3 concentrations) in asthma ED visits. The authors did not present the
results from seasonal analyses for asthma but stated that no associations were
observed between any pollutant and respiratory ED visits in the winter season.
As stated previously, in analyses of 3-h avg O3 concentrations, the authors observed
no evidence of consistent associations between any pollutant and any respiratory
outcome, including asthma. A single-city study conducted in Alberta, Canada
Villeneuve et al. (2007) from 1992-2002 among individuals two years of age and
older provides additional support for the findings from Stieb et al. (2009). but also
attempts to identify those lifestages (i.e., 2-4, 5-14, 15-44, 45-64, 65-74, or 75+) at
greatest risk to O3-induced asthma ED visits. In a time-referent case-crossover
analysis, Villeneuve et al. (2007) found an increase in asthma ED visits in an all-year
analysis across all ages (12.0% [95% CI: 6.8, 17.2] for a 30 ppb increase in max
8-h avg O3 concentrations at lag 0-2) with associations being stronger during the
warmer months (19.0% [95% CI: 11.9, 28.1]). When stratified by age, the strongest
associations were observed in the warm season for individuals 5-14 (28.1% [95% CI:
11.9, 45.1]; lag 0-2) and 15-44 (19.0% [95% CI: 8.5, 31.8]; lag 0-2). These
associations were not found to be confounded by the inclusion of aeroallergens in
age-specific models.
Several additional single-city studies have also provided evidence of an association
between asthma ED visits and ambient O3 concentrations. Ito et al. (2007b)
examined the association between short-term exposure to air pollution and asthma
ED visits for all ages in New York City from 1999 to 2002. Similar to Darrow et al.
(2011 a), when examining the spatial distribution of O3 concentrations, Ito et al.
(2007b) found a rather homogenous distribution (r > -0.80) when examining
monitor-to-monitor correlations at distances up to 20 miles. Ito et al. (2007b) used
three different weather models with varying extent of smoothing to account for
temporal relationships and multicollinearity among pollutants and meteorological
variables (i.e., temperature and dew point) to examine the effect of model selection
on the air pollutant-asthma ED visit relationship. When examining O3, the authors
reported a positive association with asthma ED visits, during the warm season across
the models (ranging from 8.6 to 16.9%) and an inverse association in the cool season
(ranging from -23.4 to -25.1%), at lag 0-1 for a 30 ppb increase in 8-h max O3
concentrations. Ito et al. (2007b) conducted copollutant models using a simplified
version of the weather model used in NMMAPS analyses (i.e., terms for same-day
temperature and 1-3 day average temperature). The authors found that O3 risk
estimates were not substantially changed in copollutant models that used every-day
data for PM2.s, NO2, SO2, and CO during the warm season (Figure 6-20 [and
Table 6-291).
6-148
-------
Ozone Warm Season
40 50 60 70
Concentration (ppb)
80
Note: The reference for the rate ratio is the estimated rate at the 5th percentile of the pollutant concentration. Estimates are
presented for the 5th percentile through the 95th percentile of pollutant concentrations due to instability in the C-R estimates at the
distribution tails.
Source: Reprinted with permission of American Thoracic Society (Strickland et al.. 2010).
Figure 6-18 Loess C-R estimates and twice-standard error estimates from
generalized additive models for associations between 8-h max
3-day average O$ concentrations and ED visits for pediatric
asthma.
Strickland et al. (2010) examined the association between O3 exposure and pediatric
asthma ED visits (ages 5-17 years) in Atlanta, GA, between 1993 and 2004 using air
quality data over the same years as Darrow et al. (201 la) and Tolbert et al. (2007).
However, unlike Darrow et al. (2011 a) and Tolbert et al. (2007), which used single
centrally located monitors or an average of monitors, respectively, Strickland et al.
(2010) used population-weighting to combine daily pollutant concentrations across
monitors. In this study, the authors developed a statistical model using hospital-
specific time-series data that is essentially equivalent to a time-stratified case-
crossover analysis (i.e., using interaction terms between year, month, and day-
of-week to mimic the approach of selecting referent days within the same month and
year as the case day). The authors observed a 6.4% (95% CI: 3.2, 9.6%) increase in
ED visits for a 30 ppb increase in 8-h max O3 concentrations at lag 0-2 in an all-year
analysis. In seasonal analyses, stronger associations were observed during the warm
season (i.e., May-October) (8.4% [95% CI: 4.4, 12.7%]; lag 0-2) than the cold season
(4.5% [95% CI: -0.82, 10.0%]; lag 0-2). Strickland et al. (2011) confirmed these
findings in an additional analysis using the same dataset, and found that the exposure
assignment approach used (i.e., centrally located monitor, unweighted average across
monitors, and population-weighted average across monitors) did not influence
6-149
-------
pediatric asthma ED visit risk estimates for spatially homogeneous pollutants such as
03.
In copollutant analyses conducted over the entire dataset for the gaseous pollutants
(i.e., CO, NO2), and limited to a subset of years (i.e., 1998-2004) for which daily PM
data (i.e., PM2.5 elemental carbon, PM2.5 sulfate) were available, Strickland et al.
(2010) found that O3 risk estimates were not substantially changed when controlling
for other pollutants (results not presented quantitatively). The authors also examined
the C-R relationship between O3 exposure and pediatric asthma ED visits and found
that both quintile and loess C-R analyses (Figure 6-18) suggest that there are elevated
associations with O3 at 8-h max concentrations as low as 30 ppb. These C-R analyses
do not provide evidence of a threshold level.
In a single-city study conducted on the West coast, Mar and Koenig (2009) examined
the association between O3 exposure and asthma ED visits (ICD-9 codes: 493-493.9)
for children (<18) and adults (> 18) in Seattle, WA from 1998 to 2002. Of the total
number of visits over the study duration, 64% of visits in the age group <18
comprised boys, and 70% of visits in the > 18 age group comprised females. Mar and
Koenig (2009) conducted a time-series analysis using both 1-h max and max 8-h avg
O3 concentrations. A similar magnitude and pattern of associations was observed at
each lag examined using both metrics. Mar and Koenig (2009) presented results for
single day lags of 0 to 5 days, but found consistent positive associations across
individual lag days which supports the findings from the studies discussed above that
examined multi-day exposures. For children, consistent positive associations were
observed across all lags, ranging from a 19.1-36.8% increase in asthma ED visits for
a 30 ppb increase in 8-h max O3 concentrations with the strongest associations
observed at lag 0 (33.1% [95% CI: 3.0, 68.5]) and lag 3 (36.8% [95% CI: 6.1, 77.2]).
Ozone was also found to be positively associated with asthma ED visits for adults at
all lags, ranging from 9.3-26.0%, except at lag 0. The slightly different lag times for
children and adults suggest that children may be more immediately responsive to O3
exposures than adults Mar and Koenig (2009).
Respiratory Infection
Although an increasing number of studies have examined the association between O3
exposure and cause-specific respiratory ED visits this trend has not included an
extensive examination of the association between O3 exposure and respiratory
infection ED visits. Stieb et al. (2009) also examined the association between short-
term O3 exposure and respiratory infection ED visits in 7 Canadian cities. In an
all-year analysis, there was no evidence of an association between O3 exposure and
respiratory infection ED visits at any lag examined (i.e., 0, 1, and 2). Across cities,
respiratory infections comprised the single largest diagnostic category,
approximately 32% of all the ED visits examined, which also included myocardial
infarction, heart failure, dysrhythmia, asthma, and COPD.
6-150
-------
6.2.7.4 Outpatient and Physician Visit Studies
Several studies have examined the association between ambient O3 concentrations
and physician or outpatient (non-hospital, non-ED) visits for acute conditions in
various geographic locations. Burra et al. (2009) examined asthma physician visits
among patients aged 1-17 and 18-64 years in Toronto, Canada from 1992 to 2001.
The authors found little or no evidence of an association between asthma physician
visits and O3; however, seasonal analyses were not conducted. It should be noted that
in this study, most of the relative risks for O3 were less than one and statistically
significant, perhaps due to an inverse correlation with another pollutant or an artifact
of the strong seasonality of asthma visits. Villeneuve et al. (2006b) also focused on
physician visits to examine the effect of short-term O3 exposure on allergic rhinitis
among individuals aged 65 or older in Toronto from 1995 to 2000. The authors did
not observe any evidence of an association between allergic rhinitis physician visits
and ambient O3 concentrations in single-day lag models in an all-year analysis
(results not presented quantitatively).
In a study conducted in Atlanta, GA, Sinclair et al. (2010) examined the association
of acute asthma and respiratory infection (e.g., upper respiratory infections and lower
respiratory infections) outpatient visits from a managed care organization with
ambient O3 concentrations as well as multiple PM size fractions and species from
August 1998 through December 2002. The authors separated the analysis into two
time periods (the first 25 months of the study period and the second 28 months of the
study period), in order to compare the air pollutant concentrations and relationships
between air pollutants and acute respiratory visits for the 25-month time-period
examined in Sinclair and Tolsma (2004) to an additional 28-month time-period of
available data from the Atlanta Aerosol Research Inhalation Epidemiology Study
(ARIES). The authors found little evidence of an association between O3 and asthma
visits, for either children or adults, or respiratory infection visits in all-year analyses
and seasonal analyses. For example, a slightly elevated RR for childhood asthma
visits was observed during the 25-month period in the cold season (RR: 1.12
[95% CI: 0.86, 1.41]; lag 0-2 for a 30 ppb increase in 8-h max O3), but not in the
warm season (RR: 0.97 [95% CI: 0.86, 1.10]; lag 0-2). During the 28-month period
at lag 0-2, a slightly larger positive effect was observed during the warm season (RR:
1.06 [95% CI: 0.97, 1.17]), compared to the cold season (RR: 1.03 [95% CI: 0.87,
1.21]). Overall, these results contradict those from Strickland et al. (2010) discussed
above. Although the mean number of asthma visits and O3 concentrations in Sinclair
et al. (2010) and Strickland et al. (2010) are similar the difference in results between
the two studies could potentially be attributed to the severity of O3-induced asthma
exacerbations (i.e., more severe symptoms requiring a visit to a hospital) and
behavior, such as delaying a visit to the doctor for less severe symptoms.
6-151
-------
6.2.7.5 Summary
The results of the recent studies evaluated largely support the conclusion of the 2006
O3 AQCD. While fewer studies were published overall since the previous review,
several multicity studies (e.g., Cakmak et al., 2006b; Dales et al., 2006) and a multi-
continent study (Katsouyanni et al., 2009) provide supporting evidence for an
association between short-term O3 exposure and an increase in respiratory-related
hospital admissions and ED visits. Across studies, different ICD-9 codes were used
to define total respiratory causes, which may contribute to some heterogeneity in the
magnitude of association. These findings are supported by single-city studies that
used different exposure assignment approaches (i.e., average of multiple monitors,
single monitor, population-weighted average) and averaging times (i.e., 1-h max and
8-h max).
Collectively, in both single-city and multicity studies there is continued evidence for
increases in both hospital admissions and ED visits when examining all respiratory
outcomes combined. Additionally, recent studies published since the 2006 O3 AQCD
support an association between short-term O3 exposure and asthma (Strickland et al..
2010: Stieb et al.. 2009) and COPD (Stieb et al.. 2009: Medina-Ramon et al.. 2006)
hospital admissions and ED visits, with more limited evidence for pneumonia-
hospital admissions and ED visits (Medina-Ramon et al.. 2006: Zanobetti and
Schwartz. 2006). As with total respiratory causes, studies used slightly different ICD-
9 codes to define specific conditions. In seasonal analyses, stronger associations were
observed in the warm season or summer months compared to the cold season,
particularly for asthma (Strickland et al.. 2010: Ito et al.. 2007b) and COPD (Medina-
Ramon et al.. 2006) (Figure 6-19 [and Table 6-281). which is consistent with the
conclusions of the 2006 O3 AQCD. There is also continued evidence that children
are particularly at greatest risk to O3-induced respiratory effects (Silverman and Ito.
2010: Strickland etal.. 2010: Mar and Koenig. 2009: Villeneuve et al.. 2007: Dales
et al.. 2006). Of note, the consistent associations observed across studies for short-
term O3 exposure and respiratory-related hospital admissions and ED visits was not
supported by studies that focused on respiratory-related outpatient or physician visits.
These differences could potentially be attributed to the severity of O3-induced
respiratory effects requiring more immediate treatment or behavioral factors that
result in delayed visits to a physician. Although the collective evidence across studies
indicates a consistent positive association between O3 exposure and respiratory-
related hospital admissions and ED visits, the magnitude of these associations may
be underestimated due to behavioral modification in response to forecasted air
quality (Neidell and Kinnev. 2010: Neidell. 2009) (Section 4.6.6).
The studies that examined the potential confounding effects of copollutants found
that O3 effect estimates remained relatively robust upon the inclusion of PM
(measured using different sampling strategies ranging from every-day to every-
6th day) and gaseous pollutants in two-pollutant models) (Figure 6-20 [and
Table 6-29]). Additional studies that conducted copollutant analyses, but did not
present quantitative results, also support these conclusions (Strickland et al.. 2010:
Tolbert et al.. 2007: Medina-Ramon et al.. 2006). Overall, recent studies provide
6-152
-------
copollutant results that are consistent with those from the studies evaluated in the
2006 O3 AQCD [(U.S. EPA. 2006bX Figure 7-12, page 7-80 of the 2006 O3 AQCD],
which found that O3 respiratory hospital admissions risk estimates remained robust
to the inclusion of PM in copollutant models.
Study
Wongetal. (2009)
Cakmak etai. 12006)
Dales etal. (2006)
Orazzoetal.(2009)a
Katsouyanni et al. (2009)
Darrowet al. (2009)
Tolbertetal. (2007)
Bigger!etal. (2005)c
Katsouyanni et al. (2009)
Stiebetal. (2009)
Villeneuyeetal. (2007)
Strickland etal. (2010)
Silverman and ltd (2010)d
Itoetal. (2007
Villeneuyeetal. (2007)
Mar and Koenig(2009
Strickland etal. (2010)
Silverman and Ito (2010)d
Mar and Koenig(2009)
Itoetal. (2007)
Villeneuyeetal. (2007)
Strickland etal. (2010)
Wongetal. (2009)
Stiebetal. (2009)
Yang etal. (2006)
Medina-Ramon etal. (2006)
Stiebetal. (2009)e
Medina-Ramon etal. (2006)
Zanobetti and Schwartz (2006)
Medina-Ramon etal. (2006)
Location
Hong Kong
10 Canadian cities
11 Canadian cities
6 Italian cities
APHENA-Europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
Atlanta
Atlanta
8 Italian cities
APHENA-Europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
7 Canadian Cities
Alberta, CAN
Atlanta
New York
New York
Alberta, CAN
Seattle, WA
Atlanta
New York
Seattle, WA
New York
Alberta, CAN
Atlanta
Hong Kong
7 Canadian Cities
Vancouver
36 U.S. cities
7 Canadian Cities
36 U.S. cities
36 U.S. cities
Boston
36 U.S. cities
36 U.S. cities
36 U.S. cities
Visit Type
HA
HA
HA
ED
HA
HA
HA
HA
ED
ED
HA
HA
HA
HA
HA
ED
ED
ED
HA
ED
ED
ED
ED
HA
ED
ED
ED
ED
HA
ED
HA
HA
ED
HA
HA
HA
HA
HA
HA
Age
All
All
0-27 days
0-2
65+
65+
65+
65+
All
All
All
65+
65+
65+
65+
All
>2
Children
All
All
>2
18+
Children
6-18
<18
All
>2
Children
All
All
65+
65+
All
65+
65+
65+
65+
65+
65+
Lag
0-1
1.2
2
0-6
0-1
0-1
DL(0-2)
DL(0-2Jb
0-2
0-3
0-1
0-1
DL(0-2)
DL(0-2Jb
2
0-2
0-2
0-1
0-1
0-2
2
0-2
0-1
0
0-1
0-2
0-2
0-1
2
0-3
DL(O-l)
NR
DL(O-l)
DL(O-I)
0-1
DL(O-l)
DL 0-1
DL 0-1
Respiratory
-25 -20 -15 -10 -5 0 5 10 15 20 25 30
% Increase
Note: Effect estimates are for a 20 ppb increase in 24-h; 30 ppb increase in 8-h max; and 40 ppb increase in 1-h max O3
concentrations. HA=hospital admission; ED=emergency department. Black=AII-year analysis; Red=Summer only analysis;
Blue=Winter only analysis.
a Wheeze used as indicator of lower respiratory disease.
bAPHENA-Canada results standardized to approximate IQR of 5.1 ppb for 1-h max O3 concentrations.
0 Study included 8 cities; but of those 8, only 4 had O3 data.
dnon-ICU effect estimates.
eThe study did not specify the lag day of the summer season estimate.
Figure 6-19 Percent increase in respiratory-related hospital admission and ED
visits in studies that presented all-year and/or seasonal results.
6-153
-------
Table 6-28 Corresponding
Study*
ED Visit or
Hospital
Admission
Effect Estimates for Figure 6-19.
Location
Age
Lag
Avg Time
% Increase
(95% Cl)
Respiratory
All-year
Wong et al. (2009)
Cakmak et al. (2006b)
Dales et al. (2006)
Orazzo et al. (2009)a
Katsouvannietal. (2009)
Hospital
Admission
Hospital
Admission
Hospital
Admission
ED Visit
Hospital
Admission
Hong Kong, China
10 Canadian cities
1 1 Canadian cities
6 Italian cities
APHENA-europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
All
All
0-27 days
0-2
65+
65+
65+
65+
0-1
1.2
2
0-6
0-1
0-1
DL(0-2)
DL(0-2)b
8-h max
24-h avg
24-h avg
8-h max
1-h max
1-h max
1-h max
1-h max
3.58(1.90,5.29)
4.38(2.19,6.46)
5.41 (2.88, 7.96)
-3.34 (-11. 2, 5.28)
1.58 (-1.71, 4.1 5)
2.38 (0.00, 4.89)
20.4 (4.07, 40.2)
2.4(0.51,4.40)
Warm
Darrow et al. (2011 a)
Tolbert et al. (2007)
Bigger! et al. (2005)°
Katsouvannietal. (2009)
ED Visit
ED Visit
Hospital
Admission
Hospital
Admission
Atlanta, GA
Atlanta, GA
8 Italian cities
APHENA-europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
All
All
All
65+
65+
65+
65+
1
0-2
0-3
0-1
0-1
DL(0-2)
DL(0-2)b
8-h max
8-h max
8-h max
1-h max
1-h max
1-h max
1-h max
2.08(1.25,2.91)
3.90 (2.70, 5.20)
0.06 (-5.24, 5.66)
0.24 (-3.32, 3.91)
2. 14 (-0.63, 4.97)
37.1 (11.5,67.5)
4.1 (1.40,6.80)
Asthma
All-year
Stieb et al. (2009)
Villeneuve et al. (2007)
Strickland et al. (2010)
ED Visit
ED Visit
ED Visit
7 Canadian cities
Alberta, CAN
Atlanta, GA
All
>2
Children
2
0-2
0-2
24-h avg
8-h max
8-h max
3.48 (0.33, 6.76)
11.9(6.8, 17.2)
6.38(3.19,9.57)
Warm
Silvermanand ltd (2010)"
Ito et al. (2007b)
Villeneuve et al. (2007)
Mar and Koenig (2009)
Strickland et al. (2010)
Silvermanand Ito (2010)"
Hospital
Admission
ED Visit
ED Visit
ED Visit
ED Visit
Hospital
Admission
New York, NY
New York, NY
Alberta, Canada
Seattle, WA
Atlanta, GA
New York, NY
All
All
>2
18+
Children
6-18
0-1
0-1
0-2
2
0-2
0-1
8-h max
8-h max
8-h max
8-h max
8-h max
8-h max
12.5(8.27,16.7)
16.9(10.9,23.4)
19.0(11.9,28.1)
19.1 (3.00,40.5)
8.43(4.42,12.7)
28.2(15.3,41.5)
Mar and Koenig (2009)
ED Visit
Seattle, WA
8-h max
33.1 (3.00, 68.5)
6-154
-------
Study*
ED Visit or
Hospital
Admission
Location Age Lag
% Increase
Avg Time (95% Cl)
Cold
Ito et al. (2007b)
Villeneuve et al. (2007)
Strickland et al. (2010)
ED Visit
ED Visit
ED Visit
New York, NY All 0-1
Alberta, Canada >2 0-2
Atlanta, GA Children 0-2
8-h max -23.4 (-27.3, -19.3)
8-hmax 8.50(0.00,17.2)
8-h max 4.52 (-0.82, 10.1)
COPD
All-year
Stieb et al. (2009)
Medina-Ramon et al. (2006)
Yang et al. (2005b)
ED Visit
Hospital
Admission
Hospital
Admission
7 Canadian cities All 2
36 U.S. cities 65+ 0-1
Vancouver, 65+ 0-3
Canada
24-h avg 4.03 (-0.54, 8.62)
8-hmax 0.24 (-0.78, 1.21)
24-h avg 8.80 (-12.5, 32.6)
Warm
Stieb et al. (2009)e
Medina-Ramon et al. (2006)
ED Visit
Hospital
Admission
7 Canadian cities All NR
36 U.S. cities 65+ 0-1
24-h avg 6.76(0.11,13.9)
8-hmax 1.63(0.48,2.85)
Cold
Medina-Ramon et al. (2006)
Hospital
Admission
36 U.S. cities 65+ 0-1
8-h max -1 .85 (-3.60, -0.06)
Pneumonia
All-year
Zanobetti and Schwartz
(2006)
Medina-Ramon et al. (2006)
Hospital
Admission
Hospital
Admission
Boston, MA 65+ 0-1
36 U.S. cities 65+ 0-1
24-h avg -5.96 (-11.1, -1 .36)
8-h max 1 .81 (-0.72, 4.52)
Warm
Medina-Ramon et al. (2006)
Hospital
Admission
36 U.S. cities 65+ 0-1
8-hmax 2.49(1.57,3.47)
Cold
Medina-Ramon et al. (2006)
Hospital
Admission
36 U.S. cities 65+ 0-1
8-hmax -4.88 (-6.59, -3.14)
'Includes studies in Fiaure 6-19.
"Wheeze used as indicator of lower respiratory disease.
bAPHENA-Canada results standardized to approximate IQR of 5.1 ppb for 1-h max O3 concentrations.
°Study included 8 cities, but of those 8 only 4 had O3 data.
dNon-ICU effect estimates.
eThe study did not specify the lag day of the summer season estimate.
6-155
-------
Study
Katsouyanni et al. (2009)a
Yangetal. (2006)a
Itoetal. (2007)b
Location Age Lag Copollutant
APHENA-U.S. 65+ 1
PM10
APHENA-Europe
PM10
APHENA-Canada
c
PM10
c PM10
Vancouver 65+ 0-3
Respiratory
COPD
•«
Asthma
New York All 0-1
CO
NO2
SO2
PM2.5
Al 1-Year
-10 -5 0
5 10 15 20 25 30
% Increase
Note: Effect estimates are for a 20 ppb increase in 24-h; 30 ppb increase in 8-h max; and 40 ppb increase in 1-h max O3
concentrations.
"Studies that examined hospital admissions.
bA study that examined ED visits.
°Risk estimates from APHENA -Canada standardized to an approximate IQR of 5.1 ppb fora 1-h max increase in O3
concentrations. Black = results from single-pollutant models; Red = results from copollutant models with PM10 or PM25;
Yellow = results from copollutant models with CO; Blue = results from copollutant models with NO2; Green = results from
copollutant models with SO2.
Figure 6-20 Percent increase in respiratory-related hospital admissions and
ED visits for studies that presented single and copollutant model
results.
6-156
-------
Table 6-29 Corresponding effect estimates for Figure 6-20.
Study*'3 Location
Visit Type Age Lag Copollutant
% Increase (95% Cl)
All-year: Respiratory
Katsouyanni APHENA-U.S.
et al. (2009)
APHENA-Europe
APHENA-
Canada
Hospital 65+ 1
Admission
PM10
PM-io
PM-io
2.62 (0.63, 4.64)
2.14 (-0.08, 4.40)
2.94(1.02,4.89)
2.38(0.08,4.64)
5.54 (-0.94, 12.4)
0.69 (-0.12, 1.50)b
5.13 (-6.62, 18.6)
0.64 (-0.87, 2.20)b
COPD
Yanq et al. Vancouver
(2005b)
Hospital 65+ 0-3
Admission
CO
NO2
SO2
PM10
8.80 (-12.5, 32.6)
22.8 (-2.14, 50.7)
11.1 (-10.4, 37.6)
13.4 (-8.40, 40.2)
11.1 (-8.40, 37.6)
Summer: Asthma
Ito et al. New York
(2007b)
'Studies included in Figure 6-20.
ED All 0-1
CO
NO2
SO2
PM2.5
r9nncn - 1-h mav Yann et al C9nnE;h^ - 94-h aurr anrl Itn et al
16.9(10.9,23.4)
18.1 (12.1,24.5)
10.2(4.29, 16.4)
13.1 (7.16, 19.5)
12.7(6.37, 19.3)
C9nn7M - R-h mav
bRisk estimates standardized to an approximate IQR of 5.1 ppb for a 1-h max increase in O3 concentrations.
To date only a few studies have examined the C-R relationship between short-term
O3 exposure and respiratory-related hospital admissions and ED visits. A preliminary
examination of the C-R relationship found no evidence of a deviation from linearity
when examining the association between short-term O3 exposure and asthma hospital
admissions (Silverman and Ito, 2010). Additionally, an examination of the C-R
relationship for O3 exposure and pediatric asthma ED visits found no evidence of a
threshold with elevated associations with O3 at concentrations as low as 30 ppb
(Silverman and Ito, 2010; Strickland et al., 2010). However, in both studies there is
uncertainty in the shape of the C-R curve at the lower end of the distribution of O3
concentrations due to the low density of data in this range.
6-157
-------
In totality, building upon the conclusions of the 2006 O3 AQCD, the evidence from
recent studies continues to support an association between short-term O3 exposure
and respiratory-related hospital admissions and ED visits. Additional evidence also
supports stronger associations during the warm season for specific respiratory
outcomes such as asthma and COPD.
6.2.8 Respiratory Mortality
The epidemiologic, controlled human exposure, and toxicological studies discussed
within this section (Section 6.2) provide evidence for multiple respiratory effects in
response to short-term O3 exposure. Additionally, the evidence from experimental
studies indicates multiple potential pathways of O3-induced respiratory effects,
which support the continuum of respiratory effects that could potentially result in
respiratory-related mortality. The 2006 O3 AQCD found inconsistent evidence for an
association between short-term O3 exposure and respiratory mortality (U.S. EPA,
2006b). Although some studies reported a strong positive association between O3
exposure and respiratory mortality, additional studies reported a small association or
no association. The majority of recent multicity studies found consistent positive
associations between short-term O3 exposure and respiratory mortality, specifically
during the summer months.
The APHENA study, described earlier in Section 6.2.7.2. (Katsouyanni et al.. 2009)
also examined associations between short-term O3 exposure and mortality and found
consistent positive associations for respiratory mortality in all-year analyses, except
in the Canadian data set for ages > 75, with an increase in the magnitude of
associations in analyses restricted to the summer season across data sets and age
ranges. Additional multicity studies from the U.S. (Zanobetti and Schwartz, 2008b),
Europe (Samoli et al., 2009), Italy (Stafoggia et al.. 2010). and Asia (Wong et al.,
2010) that conducted summer season and/or all-year analyses provide additional
support for an association between short-term O3 exposure and respiratory mortality
(Figure 6-37).
Of the studies evaluated, only the APHENA study (Katsouvanni et al.. 2009) and the
Italian multicity study (Stafoggia et al.. 2010) conducted an analysis of the potential
for copollutant confounding of the O3-respiratory mortality relationship. In the
APHENA study, specifically the European dataset, focused on the natural spline
model with 8 df/year (as discussed in Section 6.2.7.2) and lag 1 results (as discussed
in Section 6.6.2.1). respiratory mortality risk estimates were robust to the inclusion of
PMio in copollutant models in all-year analyses with O3 respiratory mortality risk
estimates increasing in the Canadian and U.S. datasets compared to single-pollutant
model results. In summer season analyses, respiratory O3 mortality risk estimates
were robust in the U.S. dataset and attenuated in the European dataset. Similarly, in
the Italian multicity study (Stafoggia et al.. 2010). which was limited to the summer
season, respiratory mortality risk estimates were attenuated in copollutant models
with PMio. Based on the APHENA and Italian multicity results, O3 respiratory
6-158
-------
mortality risk estimates appear to be moderately to substantially sensitive
(e.g., increased or attenuated) to inclusion of PMi0. However, in the APHENA study,
the mostly every-6th-day sampling schedule for PMi0 in the Canadian and U.S.
datasets greatly reduced their sample size and limits the interpretation of these
results.
6.2.9 Summary and Causal Determination
The 2006 O3 AQCD concluded that there was clear, consistent evidence of a causal
relationship between short-term O3 exposure and respiratory effects (U.S. EPA,
2006b). This conclusion was substantiated by evidence from controlled human
exposure and toxicological studies indicating a range of respiratory effects in
response to short-term O3 exposure, including pulmonary function decrements and
increases in respiratory symptoms, lung inflammation, lung permeability, and airway
hyperresponsiveness. Toxicological studies provided additional evidence for
O3-induced impairment of host defenses. Combined, these findings from
experimental studies provided support for epidemiologic evidence, in which short-
term increases in ambient O3 concentration were consistently associated with
decreases in lung function in populations with increased outdoor exposures, children
with asthma, and healthy children; increases in respiratory symptoms and asthma
medication use in children with asthma; and increases in respiratory-related hospital
admissions and asthma-related ED visits. Short-term increases in ambient O3
concentration also were consistently associated with increases in all-cause and
cardiopulmonary mortality; however, the contribution of respiratory causes to these
findings was uncertain.
Building on the large body of evidence presented in the 2006 O3 AQCD, recent
studies support associations between short-term O3 exposure and respiratory effects.
Controlled human exposure studies continue to provide the strongest evidence for
lung function decrements in young healthy adults over a range of O3 concentrations.
Studies previously reported mean O3-induced FEVi decrements of 6-8% at 80 ppb
O3 (Adams. 2006a. 2003a: McDonnell et al.. 1991: Horstman et al.. 1990). and
recent evidence additionally indicates mean FEVi decrements of 6% at 70 ppb O3
(Schelegle et al.. 2009) and 2-3% at 60 ppb O3 (Kim et al.. 2011: Brown et al.. 2008:
Adams, 2006a) (Section 6.2.1.1). In healthy young adults, O3-induced decrements in
FEVi do not appear to depend on sex (Hazucha et al., 2003), body surface area or
height (McDonnell et al., 1997), lung size or baseline FVC (Messineo and Adams,
1990). There is limited evidence that blacks may experience greater O3-induced
decrements in FEVi than do age-matched whites (Que et al., 2011; Seal et al., 1993).
Healthy children experience similar spirometric responses but lesser symptoms from
O3 exposure relative to young adults (McDonnell et al., 1985b). On average,
spirometric and symptom responses to O3 exposure appear to decline with increasing
age beyond about 18 years of age (McDonnell et al., 1999b; Seal et al., 1996). There
is also a tendency for slightly increased spirometric responses in subjects with mild
asthma and subjects with allergic rhinitis relative to healthy young adults (Torres et
6-159
-------
al., 1996). Spirometric responses in subjects with asthma appear to be affected by
baseline lung function, i.e., responses increase with disease severity (Horstman et al.,
1995).
Available information from controlled human exposure studies on recovery from O3
exposure indicates that an initial phase of recovery in healthy individuals proceeds
relatively rapidly, with acute spirometric and symptom responses resolving within
about 2 to 4 hours (Tolinsbee and Hazucha, 1989). Small residual lung function
effects are almost completely resolved within 24 hours. Effects of O3 on the small
airways persisting a day following exposure, assessed by persistent decrement in
FEF25-75% and altered ventilation distribution, may be due in part to inflammation
(Frank et al., 2001; Foster et al., 1997). In more responsive individuals, this recovery
in lung function takes longer (as much as 48 hours) to return to baseline. Some
cellular responses may not return to baseline levels in humans for more than 10-
20 days following O3 exposure (Devlin et al., 1997). Airway hyperresponsiveness
and increased epithelial permeability are also observed as late as 24 hours post-
exposure (Que et al., 2011).
With repeated O3 exposures over several days, spirometric and symptom responses
become attenuated in both healthy individuals and individuals with asthma, but this
attenuation is lost after about a week without exposure (Gong et al.. 1997a: Folinsbee
et al.. 1994: Kulle et al.. 1982). Airway responsiveness also appears to be somewhat
attenuated with repeated O3 exposures in healthy individuals, but becomes increased
in individuals with pre-existing allergic airway disease (Gong et al.. 1997a: Folinsbee
et al.. 1994). Some indicators of pulmonary inflammation are attenuated with
repeated O3 exposures. However, other markers such as epithelial integrity and
damage do not show attenuation, suggesting continued tissue damage during
repeated O3 exposure (Devlin et al., 1997).
Consistent with controlled human exposure study findings, epidemiologic evidence
indicates that lung function decrements are related to short-term increases in ambient
O3 concentration (Section 6.2.1.2). As described in the 1996 and 2006 O3 AQCDs,
the most consistent observations were those in populations engaged in outdoor
recreation, exercise, or work. Epidemiologic evidence also demonstrates that
increases in ambient O3 concentration are associated with decreases in lung function
in children with asthma (Figure 6-7 [and Table 6-81 and Figure 6-8 [and Table 6-91)
and children in the general population (Figure 6-9 [and Table 6-121). Evidence in
adults with respiratory disease and healthy adults is inconsistent. In children with
asthma, lung function mostly was found to decrease by
-------
animal studies is provided by the well-documented effects of O3 on activation of
bronchial C-fibers (Section 5.3.2).
Across disciplines, studies have examined factors that may potentially increase the
risk of Os-induced decrements in lung function. In the controlled human exposure
studies, there is a large degree of intersubject variability in lung function decrements,
symptomatic responses, pulmonary inflammation, airway hyperresponsiveness, and
altered epithelial permeability in healthy adults exposed to O3 (Que et al., 2011; Holz
et al., 2005; McDonnell, 1996). The magnitudes of pulmonary inflammation, airway
hyperresponsiveness, and increases in epithelial permeability do not appear to be
correlated, nor are these responses to O3 correlated with changes in lung function,
suggesting that different mechanisms may be responsible for these processes (Que et
al.. 2011: Balmes et al.. 1997; Balmes et al.. 1996; Arisetal.. 1995). However, these
responses tend to be reproducible within a given individual over a period of several
months indicating differences in the intrinsic responsiveness of individuals (Holz et
al.. 2005; Hazucha et al.. 2003; Holzetal.. 1999; McDonnell et al.. 1985c).
Numerous reasons for differences in the risk of individuals to O3 exposure have been
reported in the literature. These include dosimetric and mechanistic differences
(Section 5.4). Further, evidence in all three disciplines suggests a role for antioxidant
defenses (i.e., vitamin supplementation, genetic variants in oxidative metabolizing
enzymes) in modulating respiratory responses to O3. The biological plausibility of
these findings is provided by the well-characterized evidence for O3 exposure
leading to the formation of secondary oxidation products that subsequently activate
neural reflexes that mediate lung function decrements (Section 5.2.3) and that initiate
pulmonary inflammation (Section 5.3.3).
Recent controlled human exposure studies (Section 6.2.3.1) and toxicological studies
(Section 6.2.3.3) also continue to demonstrate lung injury and inflammatory
responses upon O3 exposure. Evidence from more than a hundred toxicological
studies clearly indicates that O3 induces damage and inflammation in the lung, and
studies continue to elucidate the mechanistic pathways involved (Section 5.3).
Though inflammation may resolve, continued cellular damage may alter the structure
and function of pulmonary tissues. Recent controlled human exposure studies
support previous findings for pulmonary inflammation but demonstrate effects at
60 ppb O3, the lowest concentration evaluated. Building on the extensive
experimental evidence, recent epidemiologic studies, most of which were conducted
in Mexico City, indicate ambient O3-associated increases in pulmonary inflammation
in children with asthma. Multiple studies examined and found increases in eNO
(Berhaneetal.,2011; Khatri et al.. 2009; Barraza-Villarreal et al.. 2008). In some
studies of subjects with asthma, increases in ambient O3 concentration at the same
lag were associated with both increases in pulmonary inflammation and respiratory
symptoms (Khatri et al.. 2009; Barraza-Villarreal et al.. 2008). Although more
limited in number, epidemiologic studies also found associations with cytokines such
as IL-6 or IL-8 (Barraza-Villarreal et al.. 2008; Sienra-Monge et al.. 2004).
eosinophils (Khatri et al.. 2009). antioxidants (Sienra-Monge et al.. 2004). and
indicators of oxidative stress (Romieu et al.. 2008) (Section 6.2.3.2). This
epidemiologic evidence is coherent with results from controlled human exposure and
6-161
-------
toxicological studies that demonstrated an induction or reduction of these same
endpoints after O3 exposure.
The evidence for O3-induced pulmonary inflammation and airway
hyperresponsiveness, largely demonstrated in controlled human exposure and
toxicological studies, provides mechanistic support for O3-associated increases in
respiratory symptoms observed in both controlled human exposure and
epidemiologic studies. Controlled human exposure studies of healthy, young adults
demonstrate increases in respiratory symptoms induced by O3 exposures <80 ppb
(Schelegle et al., 2009; Adams, 2006a) (Section 6.2.1.1). Adding to this evidence,
epidemiologic studies find effects in children with asthma. Evidence from the
previous large multicity NCICAS and a large body of single-city and -region studies
indicates that short-term increases in ambient O3 concentration are associated with
increases in respiratory symptoms and asthma medication use in children with
asthma (Section 6.2.4.1). Weak evidence is available from the few recent U.S.
multicity studies; however, they examined either fewer person-days of data
(Schildcrout et al., 2006) or longer lags of ambient O3 exposure (19-day avg versus
exposures lagged or averaged over a few days) than are supported by controlled
human exposure, toxicological, and other epidemiologic studies (O'Connor et al.,
2008). Several epidemiologic studies found associations between ambient O3
concentrations and respiratory symptoms in populations with asthma that also had a
high prevalence of allergy (52-100%) (Escamilla-Nufiez et al., 2008; Feo Brito et al.,
2007; Romieu et al.. 2006; Just et al.. 2002; Mortimer et al.. 2002; Ross et al.. 2002;
Gielen et al., 1997). The strong evidence in populations with asthma and allergy is
supported by observations of O3-induced inflammation in animal models of allergy
(Section 6.2.3.3), and may be explained mechanistically by the action of O3 to
sensitize bronchial smooth muscle to hyperreactivity and thus, potentially act as a
primer for subsequent exposure to antigens such as allergens (Section 5.3.5).
Modification of innate and adaptive immunity is emerging as a mechanistic pathway
contributing to the effects of O3 on asthma and allergic airways disease
(Section 5.3.6). While the majority of evidence comes from animal studies,
controlled human exposure studies have found differences between subj ects with
asthma and healthy controls in O3-mediated innate and adaptive immune responses
(Section 5.4.2.2), suggesting that these pathways may be relevant to humans and may
lead to the induction and exacerbation of asthma (Alexis et al., 2010; Hernandez et
al.. 2010; Alexis et al.. 2009; Bosson et al.. 2003).
The subclinical and overt respiratory effects observed across disciplines, as described
above, collectively provide support for epidemiologic studies that demonstrate
consistently positive associations between short-term O3 exposure and respiratory-
related hospital admissions and ED visits (Section 6.2.7). Consistent with evidence
presented in the 2006 O3 AQCD, recent multicity studies and a multicontinent study
(i.e., APHENA) (Katsouyanni et al.. 2009) found risk estimates ranging from an
approximate 1.6 to 5.4% increase in all respiratory-related hospital admissions and
ED visits in all-year analyses per unit increase in ambient O3 concentration (as
described in Section 2.5). Positive associations persisted in analyses restricted to the
6-162
-------
summer season, but the magnitude varied depending on the study location
(Figure 6-19). Compared with studies reviewed in the 2006 O3 AQCD, a larger
number of recent studies examined hospital admissions and ED visits for specific
respiratory outcomes. Although limited in number, both single- and multi-city studies
consistently found positive associations between short-term O3 exposures and asthma
and COPD hospital admissions and ED visits, with more limited evidence for
pneumonia. Consistent with the conclusions of the 2006 O3 AQCD, in studies that
conducted seasonal analyses, risk estimates were elevated in the warm season
compared to cold season or all-season analyses, specifically for asthma and COPD.
Although recent studies did not include detailed age-stratified results, the increased
risk of asthma hospital admissions (Silverman and Ito. 2010: Strickland et al.. 2010:
Dales et al.. 2006) observed for children strengthens the conclusion from the 2006 O3
AQCD that children are potentially at increased risk of O3-induced respiratory
effects (U.S. EPA. 2006b). Although the C-R relationship has not been extensively
examined, preliminary examinations found no evidence of a threshold between short-
term O3 exposure and asthma hospital admissions and pediatric asthma ED visits,
with uncertainty in the shape of the C-R curve at the lower limit of ambient
concentrations in the U.S. (Silverman and Ito. 2010: Strickland et al.. 2010).
Recent evidence extends the potential range of well-established O3-associated
respiratory effects by demonstrating associations between short-term ambient O3
exposure and respiratory-related mortality. In all-year analyses, a multicontinent
(APHENA) and multicity (PAPA) study consistently found positive associations
with respiratory mortality with evidence of an increase in the magnitude of
associations in analyses restricted to the summer months. Further, additional
multicity studies conducted in the U.S. and Europe provide evidence supporting
stronger O3-respiratory mortality associations during the summer season
(Section 6.2.8).
Several epidemiologic studies of respiratory morbidity and mortality evaluated the
potential confounding effects of copollutants, in particular, PM10, PM2.5, or NO2.
In most cases, effect estimates remained robust to the inclusion of copollutants.
In some studies of lung function and respiratory symptoms, larger effects were
estimated for O3 when copollutants were added to models. Ozone effect estimates for
respiratory-related hospital admissions and ED visits remained relatively robust upon
the inclusion of PM and gaseous pollutants in two-pollutant models (Strickland et al.,
2010: Tolbert et al., 2007: Medina-Ramon et al., 2006). Although copollutant
confounding was not extensively examined in studies of cause-specific mortality, O3-
respiratory mortality risk estimates remained positive but were moderately to
substantially sensitive (e.g., increased or attenuated) to the inclusion of PMi0 in
copollutant models (Stafoggia et al., 2010: Katsouyanni et al., 2009). However,
interpretation of these results for respiratory mortality requires caution due to the
limited PM datasets used in these studies as a result of the every 3rd- or 6th-day PM
sampling schedule employed in most cities. Together, these copollutant-adjusted
findings across respiratory endpoints provide support for the independent effects of
short-term exposures to ambient O3.
6-163
-------
Across the respiratory endpoints examined in epidemiologic studies, associations
were found using several different exposure assessment methods that likely vary in
how well ambient O3 concentrations represent ambient exposures and between-
subject variability in exposures. Evidence clearly demonstrated O3-associated lung
function decrements in populations with increased outdoor exposures for whom
ambient O3 concentrations measured on site of outdoor activity and/or at the time of
outdoor activity have been more highly correlated and similar in magnitude to
personal O3 exposures (Section 4.3.3). However, associations with respiratory effects
also were found with ambient O3 concentrations expected to have weaker personal-
ambient relationships, including those measured at home or school, measured at the
closest site, averaged from multiple community sites, and measured at a single site.
Overall, there was no clear indication that a particular method of exposure
assessment produced stronger findings.
An additional consideration in the evaluation of the epidemiologic evidence is the
impact of behavioral modifications on observed associations. A study demonstrated
that the magnitude of O3-associated asthma hospitalizations in Los Angeles, CA was
underestimated due to behavioral modification in response to forecasted air quality
(Section 4.6.5). It is important to note that the study was limited to one metropolitan
area and used air quality data for the years 1989-1997, when the O3 concentration
that determines the designation of an O3 action day, was much higher than it is
currently.
Both panel and time-series epidemiologic studies found increases in respiratory
effects in association with increases in O3 concentrations using various exposure
metrics (i.e., 24-h avg, 1-h max, and 8-h max O3 concentrations). A majority of
studies examined and found associations of respiratory symptoms with 1-h max or
8-h max O3 concentrations and associations of pulmonary inflammation with
8-h max or daytime avg O3. Within study comparisons of associations of lung
function and respiratory symptoms among various exposure metrics yielded mixed
evidence. Within some studies, larger effects were estimated for shorter O3 averaging
times whereas in other studies, larger effects were estimated for longer averaging
times or no difference was found among averaging times. Comparisons in a limited
number of time-series studies indicate rather comparable risk estimates across
exposure metrics with some evidence indicating that 24-h avg O3 was associated
with a smaller increase in risk of respiratory ED visits (Section 6.2.7.3). Overall,
there was no indication that the consistency or magnitude of the observed association
was stronger for a particular O3 exposure metric. In examination of the lag structure
of associations, epidemiologic evidence for the range of respiratory endpoints clearly
supports associations with ambient O3 concentrations lagged 0 to 1 day, which is
consistent with the O3-induced respiratory effects observed in controlled human
exposure studies. Several epidemiologic studies also found increased respiratory
morbidity in association with O3 concentrations averaged over multiple days (2 to
5 days). Across respiratory endpoints examined in epidemiologic studies, there was
not strong evidence that the magnitude of association was larger for any particular
lag.
6-164
-------
In summary, recent studies evaluated since the completion of the 2006 O3 AQCD
support and expand upon the strong body of evidence that indicated a causal
relationship between short-term O3 exposure and respiratory health effects.
Controlled human exposure studies continue to demonstrate O3-induced decreases in
FEVi and pulmonary inflammation at concentrations as low as 60 ppb.
Epidemiologic studies provide evidence that increases in ambient O3 exposure can
result in lung function decrements; increases in respiratory symptoms and pulmonary
inflammation in children with asthma; increases in respiratory-related hospital
admissions and ED visits; and increases in respiratory mortality. Recent toxicological
studies demonstrating O3-induced inflammation, airway hyperresponsiveness, and
impaired lung host defense have continued to support the biological plausibility and
modes of action for the O3-induced respiratory effects observed in the controlled
human exposure and epidemiologic studies. Additionally, recent epidemiologic
studies affirm that respiratory morbidity and mortality associations are stronger
during the warm/summer months and remain relatively robust after adjustment for
copollutants. The recent evidence integrated across toxicological, controlled human
exposure, and epidemiologic studies, along with the total body of evidence evaluated
in previous AQCDs, is sufficient to conclude that there is a causal relationship
between short-term O3 exposure and respiratory health effects.
6.3 Cardiovascular Effects
Overall, there have been a relatively small number of studies that have examined the
potential effect of short-term O3 exposure on the cardiovascular system. This was
reflected in the 1996 O3 AQCD by the limited discussion on possible O3-related
cardiovascular effects. The 2006 O3 AQCD (U.S. EPA. 2006b) built upon the limited
evidence described in the 1996 O3 AQCD (U.S. EPA. 1996a) and further explored
the potential relationship between short-term O3 exposure and cardiovascular
outcomes. The 2006 O3 AQCD concluded that "O3 directly and/or indirectly
contributes to cardiovascular-related morbidity" but added that the body of evidence
was limited. This conclusion was based on a controlled human exposure study that
included hypertensive adult males, a few epidemiologic studies of physiologic
effects, heart rate variability, arrhythmias, myocardial infarctions, and hospital
admissions, and toxicological studies of heart rate, heart rhythm, and blood pressure.
6.3.1 Controlled Human Exposure
Ozone reacts rapidly on contact with respiratory tract lining fluids and is not
absorbed or transported to extrapulmonary sites to any significant degree as such.
Controlled human exposure studies discussed in the previous AQCDs failed to
demonstrate any consistent extrapulmonary effects. Some controlled human exposure
studies have attempted to identify specific markers of exposure to O3 in blood.
Buckley et al. (1975) reported a 28% increase in serum a-tocopherol and a 26%
6-165
-------
increase in erythrocyte fragility in healthy males immediately following exposure to
500 ppb O3 for 2.75 hours with exercise (unspecified activity level). However, in
healthy adult males exposed during exercise (VE=44 L/min) to 323 ppb O3 (on
average) for 130 min on 3 consecutive days, Foster et al. (1996) found a 12%
reduction in serum a-tocopherol 20 hours after the third day of O3 exposure. Liu et
al. (1999); (1997) used a salicylate metabolite, 2,3, dehydroxybenzoic acid (DHBA),
to indicate increased levels of hydroxyl radical which hydroxylates salicylate to
DHBA. Increased DHBA levels after exposure to 120 and 400 ppb suggest that O3
increases production of hydroxyl radical. The levels of DHBA were correlated with
changes in spirometry. Interestingly, simultaneous exposure of healthy adults to O3
(120 ppb for 2 hours at rest) and concentrated ambient particles (CAPs) resulted in a
diminished systemic IL-6 response compared with exposure to CAPs alone (Urch et
al.. 2010).
Devlin et al. (2012) recently evaluated systemic and cardiovascular responses in a
group of young healthy adults (20 M, 3 F; median age 28.8 yrs) exposed to O3 (300
ppb; 2 hours with alternating 15 min periods of rest and moderate-to-heavy exercise
[VE = 25 L/min per BSA]). Relative to FA responses, immediately following the O3
exposure there was an 85% increase in blood IL-8 (p< 0.025). There were also trends
(p<0.10) for O3-induced increases in blood IL-1(3 (56%) and blood TNF-a (10%).
At 24 hrs postexposure, there were significant (p<0.025) increases blood IL-1(3
(65%) and CRP (104%). Beyond these markers of systemic inflammation, there were
also changes in biomarkers of vascular effects following O3 exposure. There were
significant (p< 0.025) O3-induced decreases in plasminogen activator inhibitor-1
(PAI-1) by 33% immediately following exposure and by 43% at 24 hrs postexposure.
Plasminogen levels were also decreased by42% at 24 hrs post exposure (p<0.05).
Finally, there was a tendency (p=0.065) for a 44% increase in tissue-type
plasminogen activator (tPA). Based on the combination of an increase in tPA and a
decrease in PAI-1, the authors suggested that O3 exposure may activate the
fibronolyisis system. Until replicated at an O3 concentration more typical of ambient
exposures, the results of this study and other high O3 concentration exposure studies
should be interpreted with caution.
Gong et al. (1998) exposed hypertensive (n = 10) and healthy (n = 6) adult males, 41
to 78 years of age, to FA and on the subsequent day to 300 ppb O3 for 3 hours with
intermittent exercise (VE = 30 L/min). The overall results did not indicate any major
acute cardiovascular effects of O3 in either the hypertensive individuals or healthy
controls. Statistically significant O3 effects for both groups combined were increases
in heart rate, rate-pressure product, and the alveolar-to-arterial PO2 gradient,
suggesting that impaired gas exchange was being compensated for by increased
myocardial work. The mechanism for the decrease in arterial oxygen tension in the
Gong et al. (1998) study could be due to an O3-induced ventilation-perfusion
mismatch. Gong et al. (1998) suggested that by impairing alveolar-arterial oxygen
transfer, the O3 exposure could potentially lead to adverse cardiac events by
decreasing oxygen supply to the myocardium. The subj ects in the Gong et al. (1998)
study had sufficient functional reserve so as to not experience significant ECG
changes or myocardial ischemia and/or injury. In studies evaluating the exercise
6-166
-------
performance of healthy adults, no significant effect of O3 on arterial O2 saturation
has been observed (Schelegle and Adams, 1986).
Fakhri et al. (2009) evaluated changes in HRV among adult volunteers (n = 50;
27 ± 7 years) during 2-hour resting exposures to PM2.5 CAPs (127 ± 62 ug/m3) and
O3 (114 ± 7 ppb), alone and in combination. High frequency HRV was increased
following CAPs-only (p = 0.046) and O3-only (p = 0.051) exposures, but not in
combination. The standard deviation of NN intervals and the square root of the mean
squared differences of successive NN intervals also showed marginally significant
(0.05< p <0.10) increase due to O3 but not CAPS. Ten of the subjects in this study
were characterized as "mildly" asthmatic, however, asthmatic status was not found to
modify these effects. Power et al. (2008) also investigated HRV in a small group of
mild-to-moderate allergic asthmatics (n = 5; mean age = 37 years) exposed for 4
hours during moderate intermittent exercise to FA, carbon and ammonium nitrate
particles (313 ± 20 ug/m3), and carbon and ammonium nitrate particles (255 ± 37
ug/m3) + O3 (200 ppb). Changes in frequency-domain variables for the particle and
particle + O3 exposures were not statistically significant compared with FA.
Seemingly in contrast to Fakhri et al. (2009), the standard deviation of NN intervals
and the square root of the mean squared differences of successive NN intervals also
showed a significant (p = 0.01) decrease for both the particle and particle + O3
exposures relative to FA responses. Using a similar protocol, Sivagangabalan et al.
(2011) concluded that spatial dispersion of cardiac repolarization was most affected
by the combined pollutant exposure of CAP + O3 compared to FA in healthy adults.
In healthy young adults (20 M, 3 F; median age 28.8 yrs), Devlin et al. (2012)
recently reported an O3-induced reduction in high frequency HRV by 51% (p<0.025)
immediately following O3 exposure (300 ppb for 2 hr with intermittent exercise) that
appeared to persist to 24 hrs postexposure (38% decrease, p<0.10). A small, 1.2%
increase in the QT interval immediately after O3 exposure relative to FA exposures
was also observed. The authors suggested that changes in HRV and repolarization
were likely mediated by nerve fibers that terminate in the lung. Changes in FEVi due
to O3 exposure are also mediated by C-fibers in the lung. There was an O3-induced
FEVi decrement of 11% in the Devlin et al. (2012) study, whereas the resting O3
exposure used by Fakhri et al. (2009) is predicted to cause very small (<0.3%)
decrements in FEVi (McDonnell et al., 2007). The induction of nerve fiber mediated
responses may, in part, explain the reduction in high frequency HRV following a
high level of exposure (300 ppb for 2 hr with intermittent exercise) in the Devlin et
al. (2012) versus the increase in high frequency HRV observed following a lower
level of exposure (120 ppb for 2hr during rest) by Fakhri et al. (2009).
Diastolic blood pressure increased by 2 mmHg following the combined O3 + CAPs
exposure, but was not altered by either O3 or CAPs alone in the Fakhri et al. (2009)
study. For a subset of the subjects without asthma in the Fakhri et al. (2009) study,
Urch et al. (2005) previously reported a 6 mmHg increase in diastolic blood pressure
following a 2-hour resting exposure to O3 (120 ppb) + PM2.5 CAPs (150 ug/m3) in
healthy adults (n = 23; 32 ± 10 years), which was statistically different from the
1 mmHg increase seen following FA exposure. Brook et al. (2002) found O3
6-167
-------
(120 ppb) + PM2.5 CAPs (150 ug/m3) in healthy adults (n = 25; 35 ± 10 years) caused
brachial artery vasoconstriction. However, minimal change in diastolic blood
pressure (0.9 mmHg increase) relative to FA (0.4 mmHg decrease) was observed.
More recently, Sivagangabalan et al. (2011) observed reported a 4.2 mmHg increase
in diastolic blood pressure following a 2-hour resting exposure to O3 (110 ppb) +
PM2.5 CAPs (150 ug/m3) in healthy adults (n = 25; 27 ± 8 years), which was
statistically different from the 1.7 mmHg increase seen following the FA exposure.
The CAP exposure alone also caused a 3 mmHg increase in diastolic blood pressure
which was significantly more than following FA. However, similar to FA, the O3
exposure alone caused a 1.8 mmHg increase in diastolic blood pressure. Overall,
these studies indicate an effect of CAPs and CAP + O3, but not O3 alone, on diastolic
blood pressure.
6.3.2 Epidemiology
The 2006 O3 AQCD concluded that the "generally limited body of evidence is highly
suggestive that O3 directly and/or indirectly contributes to cardiovascular-related
morbidity," including physiologic effects (e.g., release of platelet activating factor
[PAF]), HRV, arrhythmias, and myocardial infarctions, although the available body
of evidence reviewed during the 2006 O3 AQCD does not "fully substantiate links
between ambient O3 exposure and adverse cardiovascular outcomes" (U.S. EPA,
2006b). Since the completion of the 2006 O3 AQCD an increasing number of studies
have examined the relationship between short-term O3 exposure and cardiovascular
morbidity and mortality. These recent studies, as well as evidence from the previous
AQCDs, are presented within this section.
6.3.2.1 Arrhythmia
In the 2006 O3 AQCD, conflicting results were observed when examining the effect
of O3 on arrhythmias (Dockery et al., 2005; Rich et al., 2005). A study by Dockery et
al. (2005) reported no association between O3 concentration and ventricular
arrhythmias among patients with implantable cardioverter defibrillators (ICD) living
in Boston, MA, although when O3 concentration was categorized into quintiles, there
was weak evidence of an association with increasing O3 concentration (median O3
concentration: 22.9 ppb). Rich et al. (2005) performed a re-analysis of this cohort
using a case-crossover design and detected a positive association between O3
concentration and ventricular arrhythmias. Recent studies were conducted in various
locations and each used a different cardiac episode to define an arrhythmic event and
a different time period of exposure, which may help explain observed differences
across studies. Study-specific characteristics and air quality data for recent studies
are reported in Table 6-30.
6-168
-------
Table 6-30 Characterization of O3 concentrations (in ppb) from studies of
arrhythmias.
Study*
Metzger et al. (2007)
Rich et al. (2006b)
Location
Atlanta, GA
Boston, MA
Averaging Time
8-h max
Summer only
1-h
Mean Concentration
(Standard Deviation)
53.9 (23)
22.2*
Upper Range
of Concentration
Max: 148
75th: 33
Max: 119.5
24-h
22.6*
75th: 30.9
Max: 77.5
Rich et al. (2006a)
Anderson et al. (2010)
Sarnat et al. (2006b)
St. Louis, MO
London, England
Steubenville, OH
24-h
8-h max
24-h
Summer and Fall only
21*
8.08
21.8(12.6)
75th: 31
75th: 11.5
75th: 28.5
Max: 74.8
5 days
22.2(9.1)
75th: 29.1
Max: 44
Note: Median presented (information on mean not given); studies presented in order of first appearance in the text of this section.
Multiple studies examined O3-related effects on individuals with ICDs. A study of
518 ICD patients who had at least 1 tachyarrythmia within a 10-year period (totaling
6,287 tachyarrhythmic event-days; 1993-2002) was conducted in Atlanta, Georgia
(Metzger et al., 2007). Tachyarrhythmic events were defined as any ventricular
tachyarrhythmic event, any ventricular tachyarrhythmic event that resulted in
electrical therapy, and any ventricular tachyarrhythmic event that resulted in
defibrillation. In the primary analysis, no evidence of an association was observed for
a 30 ppb increase in 8-h max O3 concentrations and tachyarrhythmic events (OR:
1.00 [95% CI: 0.92, 1.08]; lag 0). Season-specific as well as several sensitivity
analyses (including the use of an unconstrained distributed lag model [lags 0-6])
were conducted resulting in similar null associations.
In a case-crossover analysis, a population of ICD patients in Boston, MA, previously
examined by (Rich et al., 2005) was used to assess the association between air
pollution and paroxysmal atrial fibrillation (PAF) episodes (Rich et al., 2006b).
In addition to ventricular arrhythmias, ICD devices may also detect supraventricular
arrhythmias, of which atrial fibrillation is the most common. Although atrial
fibrillation is generally not considered lethal, it has been associated with increased
premature mortality as well as hospitalization and stroke. Ninety-one
electrophysiologist-confirmed episodes of PAF were ascertained among 29 patients.
An association (OR: 3.86 [95% CI: 1.44, 10.28] per 40 ppb increase in 1-h max O3
concentrations) was observed between increases in O3 concentration during the
concurrent hour (lag 0-hour) and PAF episodes. The estimated OR for the 24-hour
moving average concentration was elevated (OR: 1.81 [95% CI: 0.86, 3.83] per
20 ppb), but weaker than the estimate for the shorter exposure window.
The association between PAF and O3 concentration in the concurrent hour during the
6-169
-------
cold months was comparable to that during the warm months. In addition, no
evidence of a deviation from linearity between O3 concentration and the log odds of
PAF was observed. Authors report that the difference between O3 concentration and
observed effect between this study (PAF and 1-hour O3) and their previous study
(ventricular arrhythmias and 24-hour moving average O3) (Rich et al.. 2005) suggest
a more rapid response to air pollution for PAF (Rich et al.. 2006b).
In an additional study, Rich et al. (2006a) employed a case-crossover design to
examine the association between air pollution and 139 confirmed ventricular
arrhythmias among 56 ICD patients in St Louis, Missouri. The authors observed a
positive association with O3 concentration (OR: 1.17 [95% CI: 0.58, 2.38] per 20 ppb
increase in 24-hour moving avg O3 concentrations [lags 0-23 hours]). Although the
authors concluded these results were similar to their results from Boston, MA (Rich
et al., 2005), they postulated that the pollutants responsible for the increased risk in
ventricular arrhythmias are different (O3 and PM2.5 in Boston and sulfur dioxide in
St Louis).
Anderson et al. (2010) used a case-crossover framework to assess air pollution and
activation of ICDs among patients from all 9 ICD clinics in the London National
Health Service hospitals. "Activation" was defined as tachycardias for which the
defibrillator delivered treatment. Investigators modeled associations using
unconstrained distributed lags from 0 to 5 days. The sample consisted of 705 patients
with 5,462 activation days (O3 concentration information was for 543 patients and
4,092 activation days). Estimates for the association with O3 concentration were
consistently positive, although weak (OR: 1.09 [95% CI: 0.76, 1.55] per 30 ppb
increase in 8-h max O3 concentrations at 0-1 day lag; OR: 1.04 [95% CI: 0.60, 1.81]
per 30 ppb increase in 8-h max O3 concentrations at 0-5 day lag) (Anderson et al..
2010).
In contrast to arrhythmia studies conducted among ICD patients, Sarnat et al. (2006b)
recruited non-smoking adults (age range: 54-90 years) to participate in a study of air
pollution and arrhythmias conducted over two 12-week periods during summer and
fall of 2000 in a region characterized by industrial pollution (Steubenville, Ohio).
Continuous ECG data acquired on a weekly basis over a 30-minute sampling period
were used to assess ectopy, defined as extra cardiac depolarizations within the atria
(supraventricular ectopy, SVE) or the ventricles (ventricular ectopy, VE). Increases
in the 5-day moving average (days 1-5) of O3 concentration were associated with an
increased odds of SVE (OR: 2.17 [95% CI: 0.93, 5.07] per 20 ppb increase in
24-h avg O3 concentrations). A weaker association was observed for VE (OR: 1.62
[95% CI: 0.54, 4.90] per 20 ppb increase in 24-h avg O3 concentrations). The results
of the effect of 5-day O3 concentration on SVE were robust to the inclusion of SO42"
in the model [OR: 1.62 (95% CI: 0.54, 4.90)]. The authors indicate that the strong
associations observed at the 5-day moving averages, as compared to shorter time
periods, suggests a relatively long-acting mechanistic pathways, such as
inflammation, may have promoted the ectopic beats in this population (Sarnat et al..
2006b).
6-170
-------
Although many studies report positive associations, collectively, studies of
arrhythmias report inconsistent results. This may be due to variation in study
populations, length and season of averaging time, and outcome under study.
6.3.2.2 Heart Rate/Heart Rate Variability
In the 2006 O3 AQCD, two large population-based studies of air pollution and HRV
were summarized (Park et al., 2005b; Liao et al., 2004a). In addition, the biological
mechanisms and potential importance of HRV were discussed. Briefly, the study of
acute effects of air pollution on cardiac autonomic control is based on the hypothesis
that increased air pollution levels may stimulate the autonomic nervous system and
lead to an imbalance of cardiac autonomic control characterized by sympathetic
activation unopposed by parasympathetic control (U.S. EPA, 2006b). Examples of
HRV indices include the standard deviation of normal-to-normal intervals (SDNN),
the square root of the mean of the sum of the squares of differences between adjacent
NN intervals (r-MSSD), high-frequency power (HF), low-frequency power (LF), and
the LF/HF ratio. Liao et al. (2004a) examined the association between air pollution
and cardiac autonomic control in the fourth cohort examination (1996-1998) of the
U.S.-based Atherosclerosis Risk in Communities Study. A decrease in log-
transformed HF was associated with an increase in O3 concentration among white
study participants. Park et al. (2005b) examined the effects of air pollution on indices
of HRV in a population-based study among men from the Normative Aging Study in
Boston, Massachusetts. Several associations were observed with O3 concentration
and HRV outcomes. A reduction in LF was associated with increased O3
concentration, which was robust to inclusion of PM2.s. The associations with all
HRV indices and O3 concentration were stronger among those with ischemic heart
disease and hypertension. In addition to the population-based studies included in the
2006 O3 AQCD was a study by Schwartz et al. (2005). who conducted a panel study
to assess the relationship between exposure to summertime air pollution and HRV.
A weak association of O3 concentration during the hour immediately preceding the
health measures was observed with r-MSSD among a study population that consisted
of mostly older female participants. In summary, these studies suggest that short-
term exposures to ambient O3 concentrations are predictors of decreased HRV and
that the relationship may be stronger among certain subgroups. More recent studies
that examined the association between O3 concentration and HRV are described
below. Study-specific characteristics and O3 concentrations for these studies are
presented in Table 6-31.
6-171
-------
Table 6-31 Characterization of O3 concentrations (in ppb) from studies of heart
rate variability.
Study*
Parketal. (2007)
Parketal. (2008)
Baiaetal. (2010)
Wheeler et al. (2006)
Zanobetti et al. (2010)
Ruidavets et al. (2005a)
Hampeletal. (2012)
Chan et al. (2005a)
Wuetal. (2010)
Chuang et al. (2007a)
Chuang et al. (2007b)
Location
Boston, MA
Boston, MA
Boston, MA
Atlanta, GA
Boston, MA
Toulouse, France
Augsburg, Germany
Taipei, Taiwan
Taipei, Taiwan
Taipei, Taiwan
Taipei, Taiwan
Averaging Time
24-h
24-h
Olag
1 0-h lag
4-h
24-h
0.5-h
2-h
3-D
5-D
8-h max
1 -h avg
1-h
Working period
24-h
48-h
72-h
1-h
Mean Concentration
(Standard Deviation)
Range of 17.0-29.1
23.4(13)
23 (1 6)
21 (15)
18.5
29.4
20.7*
20.5*
21.9*
22.8*
38.3(14.8)
23.4(17.0)
21.9(15.4)
24.9(14.0)
28.4(12.1)
33.3 (8.9)
33.8(7.1)
35.1
Upper Range of
Concentration
75th: 22.5
75th: 30.33
75th: 30.08
75th: 28.33
75th: 29.28
75th: 46.9
Max: 80.3
75th: 35.2
Max: 80.6
Max: 114.9
Max: 59.2
Max: 49.3
Max: 47.8
Max: 48.3
Max: 192.0
*Note: Median presented (information on mean not given); studies presented in order of first appearance in the text of this section.
Several follow-up examinations of HRV were conducted among the participants of
the Normative Aging Study in Boston, Massachusetts. A trajectory cluster analysis
was used to assess whether pollution originating from different locations had varying
relationships with HRV (Park et al., 2007). Subjects who were examined on days
when air parcels originated in the west had the strongest associations with O3;
however, the O3 concentration in this cluster was low (24-h avg, 17.0 ppb) compared
to the other clusters (24-h avg of 21.3-29.1 ppb). LF and SDNN decreased with
increases in the 4-hour moving average of O3 concentration from the west (LF
decreased by 51.2% [95% CI: 1.6, 75.9%] and SDNN decreased by 28.2% [95% CI:
-0.5, 48.7%] per 30 ppb increase in 4-h avg O3 concentrations) (Park et al.. 2007).
The Boston air mass originating in the west traveled over Illinois, Indiana, and Ohio;
states typically characterized by coal-burning power plants. Due to the low O3
concentrations observed in the west cluster, the authors hypothesize that O3
concentration on those days could be capturing the effects of other, secondary and/or
transported pollutants from the coal belt or that the relationship between ambient O3
concentration and personal exposure to O3 is stronger during that period (supported
by a comparatively low apparent temperature which could indicate a likelihood to
keep windows open and reduced air conditioning use) (Park et al.. 2007).
6-172
-------
An additional follow-up evaluation using the Normative Aging Study examined the
potential for effect modification by chronic lead (Pb) exposure on the relationship
between air pollution and HRV (Park et al., 2008). Authors observed graded
reductions in HF and LF of HRV in relation to O3 (and sulfate) concentrations across
increasing quartiles of tibia and patella lead (HF: percent change 32.3% [95% CI:
-32.5, 159.3] for the first quartile of tibia Pb and -59.1 [95% CI: -77.3, -26.1] for the
fourth quartile of tibia Pb per 30 ppb increase in 4-h avg O3 concentrations; LF:
percent change 8.0% [95% CI: -36.9, 84.9] for the first quartile of tibia Pb and -59.3
[95% CI: -74.6, -34.8] for the fourth quartile of tibia Pb per 30 ppb increase in
4-h avg O3 concentrations). In addition, associations were similar when education
and cumulative traffic-adjusted bone Pb levels were used in analyses. Authors
indicate the possibility that O3 (which has low indoor concentrations) was acting as a
proxy for sulfate (correlation coefficient for O3 and sulfate = 0.57). Investigators of a
more recent follow-up to the Normative Aging Study hypothesized that the
relationships between short-term air pollution exposures and ventricular
repolarization, as measured by changes in the heart-rate corrected QT interval (QTc),
would be modified by participant characteristics (e.g., obesity, diabetes, smoking
history) and genetic susceptibility to oxidative stress (Bajaet al.. 2010). No evidence
of an association between O3 concentration (using a quadratic constrained distributed
lag model and hourly exposure lag models over a 10-hour time window preceding
the visit) and QTc was reported (change in mean QTc -0.74 [95% CI: -3.73, 2.25]);
therefore, potential effect modification of personal and genetic characteristics with
O3 concentration was not assessed (Baja et al.. 2010). Collectively, the results from
studies that examined the Normative Aging Study cohort found an association
between increases in short-term O3 concentration and decreases in HRV (Park et al..
2008; Park et al., 2007; Park et al., 2005b) although not consistently in all of the
studies (Baja et al., 2010). Further, observed relationships appear to be stronger
among those with ischemic heart disease, hypertension, and elevated bone lead
levels, as well as when air masses arrive from the west (the coal belt). However, it is
not clear if O3 concentration is acting as a proxy for other, secondary particle
pollutants (such as sulfate) (Park et al., 2008). In addition, since the Normative
Aging Study participants were older, predominately white men, results may not be
generalizable to the a large proportion of the U.S. population.
Additional studies of populations not limited to the Normative Aging Study have also
examined associations between O3 exposure and HRV. A panel study among 18
individuals with COPD and 12 individuals with recent myocardial infarction (MI)
was conducted in Atlanta, Georgia (Wheeler et al.. 2006). HRV was assessed for
each participant on 7 days in fall 1999 and/or spring 2000. Ozone concentrations
were not associated with HRV (SDNN) among all subjects (percent change of 2.36%
[95% CI: -10.8%, 17.5%] per 30 ppb 4-hour O3 increase) or when stratified by
disease type (COPD, recent MI, and baseline FEVi) (Wheeler et al.. 2006).
HRV and air pollution was assessed in a panel study among 46 predominately white
male patients (study population: 80.4% male, 93.5% white) aged 43-75 years in
Boston, Massachusetts, with coronary artery disease (Zanobetti et al., 2010). Up to
four home visits were made to assess HRV over the year following the index event.
6-173
-------
Pollution lags used in analyses ranged between 30 minutes to a few hours and up to
5 days prior to the HRV assessments, calculated from hourly O3 measurements
averaged over three monitoring sites in Boston. Decreases in r-MSSD were reported
for all averaging times of O3 concentration (percent change of-5.18% [95% CI:
-7.89, -2.30] per 20 ppb of 5-day moving average of O3 concentration), but no
evidence of an association between O3 concentration and HF was observed
(quantitative results not provided). In two-pollutant models with O3 and either PM2.5
or BC, O3 associations remained robust.
A few recent studies were conducted outside of the U.S. in Europe (Hampel et al.,
2012: Ruidavets et al.. 2005a) and Asia (Wu et al.. 2010: Chuang et al.. 2007b:
Chuang et al., 2007a: Chan et al., 2005a: Ruidavets et al., 2005a) that also examined
the relationship between air pollution concentrations and heart rate and HRV.
No consistent relationships were identified between O3 concentration and resting
heart rate among middle-aged (35-64 years) participants residing in Toulouse, France
(Ruidavets et al., 2005a). A negative trend was reported for the 3-day cumulative
(lag days 1-3) concentration of 8-h max O3 with heart rate (p for trend = 0.02);
however, the individual odds ratios comparing quintiles of exposure showed no
association (OR for O3 concentraction of 0.93 [95% CI: 0.86, 1.01] for the highest
quintile of resting heart rate compared to the lowest). When stratified by current
smoking status, non-smokers had a decreased trend with increased 3-day cumulative
O3 concentrations but none of the quintiles for heart rate were statistically
significant. In a panel study conducted in Augsburg, Germany, Hampel et al. (2012)
examined the effect of short-term O3 exposures on measures of heart rate and
repolarization in individuals with type 2 diabetes or impaired glucose tolerance and
healthy individuals with a potential genetic predisposition. A ~1% increase in HR
was observed for individuals with type 2 diabetes and impaired glucose tolerance at
concurrent and lag 1-4 hours for an approximate 10 ppb increase in O3
concentrations1; no effect was observed for healthy individuals. These associations
remained robust in copollutants models with sulfate, PM, and ultrafme particles.
Additionally, there was evidence of T-wave flattening across all lags in healthy
individuals and those with type 2 diabetes and impaired glucose intolerance, with the
effect strongest in these individuals at concurrent (-1.31% [95% CI: -2.19, -0.42])
and lag 1 hour (-1.32% [95% CI: -2.19, 0.45]). Similarly, there was evidence of an
increase in T-wave complexity for all participants across all lags examined, with the
strongest effects again for individuals with type 2 diabetes and impaired glucose
tolerance, but at lags 1 and 2 hours. An increase in T wave complexity for healthy
participants at lags of 3 and 4 hours. Hampel et al. (2012) also found evidence of
effect modification for each of the heart rate and repolarization metrics when taking
into consideration the location and season in which the ECG recordings were
obtained, with greater effects occurring when measurements were taken outdoors
during the summer.
1 These results were not standardized to a 1 -h max O3 concentration of 40 ppb because the study examined hourly changes in heart
rate parameters. Using an increment of 40 ppb would not appropriately represent the potential hourly change in O3
concentrations.
6-174
-------
In a study conducted in Taipei, Taiwan no associations were reported between O3
concentration and HRV among CHD patients and patients with one or more major
CHD risk factors (Chan et al., 2005a). Another study in Taipei, Taiwan examined
mail carriers and reported O3 concentration measured using personal monitors.
No association was observed between O3 concentration and the measures of HRV
(percent change for SDNN: 0.57 [95% CI: -21.27, 28.46], r-MSSD: -7.10 [95% CI:
-24.24, 13.92], HF: -1.92 [95% CI: -23.68, 26.02], LF: -4.82 [95% CI: -25.34, 21.35]
per 40 ppb O3) (Wu et al.. 2010). A panel study was conducted in Taiwan to assess
the relationship between air pollutants and inflammation, oxidative stress, blood
coagulation, and autonomic dysfunction (Chuang et al.. 2007b: Chuang et al..
2007a). Participants were apparently healthy college students (aged 18-25 year) who
were living in a university dormitory in metropolitan Taipei. Health endpoints were
measured three times from April to June in 2004 or 2005. Ozone concentration was
assessed in statistical models using the average of the 24, 48, and 72 hours before the
hour of each blood sampling. Decreases in HRV (measured as SDNN, r-MSSD, LF,
and HF) were associated with increases in O3 concentrations in single-pollutant
models (percent change for SDNN: -13.45 [95% CI: -16.26, -10.60], r-MSSD -13.76
[95% CI: -21.62, -5.44], LF -9.16 [95% CI: -13.29, -4.95], HF -10.76 [95% CI:
-18.88, -2.32] per 20 ppb cumulative 3-day avg O3 concentrations) and remained
associated with 3-day O3 concentrations in two-pollutant models with sulfate.
Another study in Taiwan recruited individuals with CHD or at risk for cardiovascular
disease from outpatient clinics during the study period (two weeks in February)
(Chuang et al.. 2007b). No association was observed between O3 concentration and
HRV measures (SDNN, r-MSSD, LF, HF) (numerical results not provided in
publication).
Overall, studies of O3 concentration and HRV report inconsistent results. Multiple
studies conducted in Boston, MA, observed positive associations but the authors of
many of these studies postulated that O3 concentration was possibly acting as a proxy
for other pollutants. The majority of other studies, both in the U.S. and
internationally, report null findings. The inconsistencies observed are further
complicated by the different HRV measures and averaging times used by the studies.
6.3.2.3 Stroke
The 2006 O3 AQCD did not identify any studies that examined the association
between short-term O3 exposure and stroke. However, recent studies have attempted
to examine this relationship. Lisabeth et al. (2008) used a time-series approach to
assess the relationship between daily counts of ischemic stroke and transient
ischemic attack (TIA) with O3 concentrations in a southeast Texas community
among residents 45 years and older (2001-2005; median age of cases, 72 years).
The median O3 concentration (hourly average per 24-hour time-period) was 25.6 ppb
(IQR 18.1-33.8). The associations between same-day O3 concentrations and
stroke/TIA risk were positive (RR: 1.03 [95% CI: 0.96, 1.10] per 20 ppb increase in
24-h avg O3 concentrations) and previous-day (RR: 1.05 [95% CI: 0.99, 1.12] per
6-175
-------
20 ppb increase in 24-h avg O3 concentrations). Associations were robust to
adjustment for PM2.5.
A case-crossover design was used in a study conducted in Dijon, France between
March 1994 and December 2004, among those 40 years of age and older who
presented with first-ever stroke (Henrotin et al. 2007). The mean O3 concentration,
calculated over 8-hour daytime periods, was 14.95 ppb (IQR: 6-22 ppb).
No association was observed between O3 concentration at any of the single-day lags
examined (i.e., 0-3 days) and hemorrhagic stroke. However, an association between
ischemic stroke occurrence and O3 concentrations with a 1-day lag was observed
(OR: 1.54 [95% CI: 1.14, 2.09] per 30 ppb increase in 8-h max O3 concentrations).
The observed association between short-term O3 exposure and ischemic stroke
persisted in two-pollutant models with PMi0, SO2, NO2, or CO. This association was
stronger among men (OR: 2.12 [95% CI: 1.36, 3.30] per 30 ppb increase in 8-h max
O3 concentrations) than among women (OR: 1.17 [95% CI: 0.77, 1.78] per 30 ppb
increase in 8-h max O3 concentrations) in single pollutant models. When stroke was
examined by subtype among men, an association was observed for ischemic strokes
of large arteries and for transient ischemic attacks, but not for cardioembolic or
lacunar ischemic strokes. The subtype analysis was not performed for women.
Additionally, for men a linear exposure-response was observed when O3
concentration was assessed based on quintiles (p for trend = 0.01) (Figure 6-21).
A potential limitation of this study is that 67.4% of the participating men were
smokers compared to 9.3% of the women.
Another case-crossover study performed in Dijon, France examined the association
between O3 concentration and incidence of fatal and non-fatal ischemic
cerebrovascular events (ICVE) (Henrotin et al.. 2010). Mean 8-hour O3
concentration was 19.1 ppb (SD 12.2 ppb). A positive association was observed
between recurrent ICVE and 8-hour O3 concentration with a 3-day lag (OR: 1.92
[95% CI 1.17, 3.12]), but not for other lags (0, 1, 2, 4) or cumulative days (0-1, 0-2,
1-2, 1-3). Although some ORs for incident ICVEs were elevated, none were
statistically significant. Results for associations using the maximum daily 1-hour O3
concentrations were similar to the 8-hour results but slightly attenuated. ORs were
similar in two pollutant models with SO2, NO2, CO, and PM10 (data not given).
In stratified analyses, the association between 1-day lagged O3 concentration and
incident and recurrent ICVE was greater among individuals with diabetes or
individuals with multiple pre-existing vascular conditions.
6-176
-------
3.5
3
2.5-
i *
1.5 -
•o
8
1 -
0.5-
0-8 9-20 21-32 33-48
O3 concentration (ppb)
48-115
Source: Henrotin et al. (2007).
Figure 6-21 Odds ratio (95% confidence interval) for ischemic stroke by
quintiles of O$ exposure.
6.3.2.4 Biomarkers
An increasing number of studies have examined the relationship between air
pollution and biomarkers in an attempt to elucidate the biological mechanisms
linking air pollution and cardiovascular disease. A wide range of markers assessed as
well as different types of study designs and locations chosen make comparisons
across studies difficult. Table 6-32 provides an overview of the O3 concentrations
reported in each of the studies evaluated.
6-177
-------
Table 6-32 Characterization of O3 concentrations (in ppb) from studies of
biomarkers.
Study*
Liao et al. (2005)
Thompson et al.
(2010)
Rudez et al. (2009)
Chuana et al.
(2007a)
Steinvil et al. (2008)
Chen et al. (2007a)
Wellenius et al.
(2007)
Goldberg et al.
(2008)
Baccarelli et al.
(2007)
Chuana et al.
(2010)
Location
3 U.S. counties
Toronto, Ontario
Rotterdam, the Netherlands
Taipei, Taiwan
Tel-Aviv, Israel
Los Angeles and
San Francisco, CA
Boston, MA
Montreal, Quebec
Lombardia, Italy
Taiwan
Averaging
Time
8-h
1-h/1 yr
24-h
24-h
48-h
72-h
0.5-h
8-h / 2 weeks
8-h / 1 mo
1-h/24-h
24-h
1-h
Mean
Concentration
(Standard
Deviation)
40 (20)
21.94(15.78)
22*
28.4(12.1)
33.3 (8.9)
33.8(7.1)
29.2 (9.7)
30.8*
28.3*
25.1 (12.9)
NS
18.3*
26.83 (9.7)
Upper Range of
Concentration
75th: 31 .5
Max: 90
Max: 49.3
Max: 47.8
Max: 48.3
75th: 36
Max: 47.9
Max: 43.1
75th: 35.1
Max: 202.3
Max: 62.1
*Note: Median presented (information on mean not given); studies presented in order of first appearance in the text of this section.
Hemostasis and coagulation markers
Multiple studies used various markers to examine if associations were present
between short-term O3 exposure and hemostasis and coagulation. Some of the
markers included in these studies were as follows: fibrinogen, von Willebrand factor
(vWF), plasminogen activator fibrinogen inhibitor-1 (PAI-1), tissue-type
plasminogen activator (tPA), platelet aggregation, and thrombin generation.
A population-based study in the United States was conducted to assess the
relationship between short-term exposure to air pollution and markers of blood
coagulation using the Atherosclerosis Risk in Communities (ARIC) study cohort
(Liao et al., 2005). Significant curvilinear associations were observed for O3 (1 day
prior to blood draw) and fibrinogen and vWF (quantitative results not provided for
regression models although adjusted means [SE] of vWF were given as 118%
[0.79%] for O3 concentrations <40 ppb, 117% [0.86%] for O3 concentrations 40-
70 ppb, and 124% [1.97%] for O3 concentrations of 70 ppb). The association
between short-term O3 exposure and fibrinogen was more pronounced among those
with a history of cardiovascular disease (CVD) and was statistically significant
6-178
-------
among only this subgroup of the population. The curvilinear relationship between
concentration and outcome suggested stronger relationships at higher concentrations
of O3. The authors note that the most pronounced associations occurred when the
pollutant concentrations were 2-3 standard deviations above the mean. The results
from this relatively large-scale cross-sectional study suggest weak associations with
between short-term O3 exposure and increases in fibrinogen (among those with a
history of CVD) and vWF. A retrospective repeated measures analysis was
performed in Toronto, Canada among adults aged 18-40 years (n = 45) between the
years of 1999 and 2006 (Thompson et al.. 2010). Single pollutant models were used
with moving averages up to 7 days. No evidence of an association was observed
between short-term O3 exposure and increases in fibrinogen.
A repeated measures study was conducted among 40 healthy individuals living or
working in the city center of Rotterdam, the Netherlands to assess the relationship
between air pollution and markers of hemostatis and coagulation (platelet
aggregation, thrombin generation, and fibrinogen) (Rudez et al., 2009). Each
participant provided between 11 and 13 blood samples throughout a 1-year period
(498 samples on 197 days). Examined lags ranged from 6 hours to 3 days prior to
blood sampling. No consistent evidence of an association was observed between O3
concentration and any of the biomarkers (percent change of max platelet aggregation:
-6.87 [95% CI: -21.46, 7.70] per 20 ppb increase in 24-h avg O3 concentration at
4-day average; percent change of endogenous thrombin potential: 0.95 [95% CI:
-3.05, 4.95] per 20 ppb increase in 24-h avg O3 concentration at 4-day avg; percent
change of fibrinogen: -0.57 [95% CI: -3.05, 2.00] per 20 ppb increase in 24-h avg O3
concentration at lag 1-day). Some associations with O3 were in the opposite direction
to that hypothesized which may be explained by the negative correlation between O3
and other pollutants (correlation coefficients ranged from -0.4 to -0.6).
The statistically significant inverse effects observed in single-pollutant models with
O3 were no longer apparent when PMi0 was included in the model (Rudez et al..
2009).
A panel study in Taiwan measured health endpoints using blood samples from
healthy individuals (n = 76) at three times from April to June in 2004 or 2005
(Chuang et al., 2007a). Increases in fibrinogen and PAI-1 were associated with
increases in O3 concentrations in single-pollutant models (percent change in
fibrinogen: 11.76 [95% CI: 4.03, 19.71] per 20 ppb 3-day cumulative avg O3
concentration; percent change in PAI-1: 6.08 [95% CI: 38.91, 84.27] per 20 ppb
3-day cumulative avg O3 concentration). These associations were also observed at 1
and 2 day averaging times. Associations between PAI-1 and 3-day O3 concentrations
remained robust in two-pollutant models with sulfate. No association was observed
between O3 concentration and tPA, a fibrinolytic factor (percent change 16.15
[95% CI: -4.62, 38.34] per 20 ppb 3-day avg O3 concentration).
A study in Israel examined the association between pollutant concentrations and
fibrinogen among 3,659 apparently healthy individuals (Steinvil et al.. 2008).
In single pollutant models, O3 was associated with an increase in fibrinogen at a
4-day lag among men and a same-day O3 concentration among women but results for
6-179
-------
other lags (0 through 7 days) were mixed (i.e., some positive and some negative;
none statistically significant).
Inflammatory markers
Potential associations between short-term exposures to air pollution and
inflammatory markers (C-reactive protein [CRP], white blood cell [WBC] count,
albumin, and Interleukin-6 [IL-6]) were also examined in several studies.
The ARIC study cohort, which included men and women aged 45-64 years old at the
start of the study, was utilized to assess the association between O3 concentrations
and markers of inflammation, albumin and WBC count (Liao et al.. 2005).
No association was observed between O3 concentrations and albumin or WBC count.
Thompson et al. (2010) assessed ambient air pollution exposures and IL-6. This
retrospective repeated measures analysis was conducted among 45 adults (18-
40 years of age) in Toronto, Canada between the years of 1999 and 2006. Single
pollutant models were used to analyze the repeated-measures data using moving
averages up to 7 days. A positive association was observed between IL-6 and short-
term 1-hour O3 exposure with the strongest effects observed for the average of lags
0-3 days (quantitative results not provided). No association was observed for shorter
averaging times (average lags of <1 day). When examined by season using 2-day
moving averages, the association between short-term O3 exposure and IL-6 was
positive during only the spring and summer.
In Rotterdam, the Netherlands, a repeated measures study of healthy individuals
living or working in the city center reported no association between short-term O3
exposure and CRP (Rudez et al., 2009). Each of the 40 participants provided between
11 and 13 blood samples throughout a 1-year period (498 samples on 197 days).
No consistent evidence of an association was observed between O3 concentration and
CRP (percent change: -0.48 [95% CI: -14.05, 13.10] per 20 ppb increase in 24-h avg
O3 concentration at lag 1-day). Additionally, no association was observed with 2 or
3 day lags.
The relationship between pollutant concentrations and one-time measures of
inflammatory biomarkers was assessed in sex-stratified analyses among 3,659
apparently healthy individuals in Tel Aviv, Israel (Steinvil et al.. 2008). No evidence
of an association was observed between O3 concentration and CRP or WBC for men
and women.
A panel study of healthy individuals (n = 76) was conducted in Taiwan to assess the
relationship between air pollutants and inflammation (Chuang et al., 2007a). Health
endpoints were measured three times from April to June in 2004 or 2005. Ozone
effects were assessed in statistical models using the average of the 24 hours (1 day),
48 hours (2 days), and 72 hours (3 days) before the hour of each blood sampling.
Increases in CRP were associated with increases in O3 concentrations in single-
pollutant models (percent change in CRP: 244.38 [95% CI: 4.54, 585.15] per 20 ppb
6-180
-------
3-day avg O3 concentration). The association was also observed using a 2-day
cumulative averaging time, but no association was present with a 1-day averaging
time.
Oxidative stress markers
A few studies have reported on the relationships between short-term O3 exposure and
increases in markers of oxidative stress. The association between O3 concentration
and markers of lipid peroxidation and antioxidant capacity was examined among 120
nonsmoking healthy college students, aged 18-22 years, from the University of
California, Berkeley (February-June 2002) (Chen et al., 2007a). By design, students
were chosen that had experienced different geographic concentrations of O3 over
their lifetimes and during recent summer vacation in either greater Los Angeles (LA)
or the San Francisco Bay Area (SF). Long-term (based on lifetime residential
history) and shorter-term (based on the moving averages of 8-h max concentrations
1-30 days prior to the day of blood collection) O3 concentration were estimated
(lifetime exposure results are presented in Chapter 7). A marker of lipid peroxidation,
8-isoprostane (8-iso-PGF), was assessed. This marker is formed continuously under
normal physiological conditions but has been found at elevated concentrations in
response to environmental exposures. A marker of overall antioxidant capacity, ferric
reducing ability of plasma (FRAP), was also measured. Levels of 8-iso-PGF were
associated with 2-week ((3 = 0.035 [pg/mL]/8-hour ppb O3, p = 0.007) and 1-month
(P = 0.031 [pg/mL]/8-hour ppb O3, p = 0.006) estimated O3 concentrations.
No evidence of association was observed between short-term O3 exposure and
increases in FRAP. A chamber study performed among a subset of study participants
supported the primary study results. The concentrations of 8-iso-PGF increased
immediately after the 4-hour controlled O3 exposure ended (p = 0.10). However,
levels returned to near baseline by 18 hours without further exposure. The authors
note that O3 was highly correlated with PMio_2.s and NO2 in this study population;
however, O3 associations remained robust in copollutant models.
Using blood samples collected between April and June of 2004 or 2005 in Taiwan,
the association between short-term O3 exposure and a marker of oxidative stress
(i.e., 8-hydroxy-2'-deoxyguanosine (8-OHdG)) was studied among healthy
individuals (n = 76) (Chuang et al., 2007a). Increases in 8-OHdG were associated
with increases in O3 concentrations in single-pollutant models (percent change in 8-
OHdG: 2.46 [95% CI: 1.01, 3.92] per 20 ppb increase in 24-h avg O3).
The association did not persist with 2- or 3-day cumulative averaging times.
Markers of overall cardiovascular health
Multiple studies used markers that assess overall cardiovascular well-being.
Wellenius et al. (2007) examined B-type natriuretic peptide (BNP), a marker of heart
failure, in a repeated-measures study conducted in Boston, MA, among 28 patients
with congestive heart failure and impaired systolic function. The authors found no
6-181
-------
evidence of an association between BNP and short-term O3 exposures at lags 0-
3 days (quantitative results not provided). BNP was chosen because it is directly
associated with cardiac hemodynamics and symptom severity among those with heart
failure and is considered a marker of functional status. However, the authors
conclude that the use of BNP may not be useful in studies of the health effects of
ambient air pollutants due to the large amount of within-person variability in BNP
levels observed in this population.
The relationship between air pollution and oxygen saturation and pulse rate, markers
of physiological well-being, was examined in a 2-month panel study among 31
congestive heart failure patients (aged 50-85 years) in Montreal, Canada from July
2002 to October 2003 (Goldberg et al., 2008). All participants had limited physical
functioning (New York Heart Association Classification > II) and an ejection fraction
(the fraction of blood pumped out of the heart per beat) less than or equal to 35%
(normal is above 55%). Daily mean O3 concentrations were calculated based on
hourly measures at 10 monitoring stations. There was an inverse association between
O3 concentration (lag-0) and oxygen saturation when adjustment was made for
temporal trends. In the models incorporating personal covariates and weather factors,
the association remained but was not statistically significant. The associations of O3
concentration with a lag of 1 day or a 3-day mean were not statistically significant.
No evidence of association was observed between O3 concentration and pulse rate.
Total homocysteine (tHcy) is an independent risk factor for vascular disease and
measurement of this marker after oral methionine load is used to identify individuals
with mild impairment of homocysteine metabolism. The effects of air pollution on
fasting and postmethionine-load tHcy levels were assessed among 1,213 apparently
healthy individuals from Lombardia, Italy from January 1995 to September 2005
(Baccarelli et al., 2007). A 20-ppb increase in the 24-h avg O3 concentrations was
associated with an increase in fasting tHcy (percent change 6.25 [95% CI: 0.84,
11.91]) but no association was observed with postmethionine-load tHcy (percent
change 3.36 [95% CI: -1.30, 8.39]). In addition, no evidence of an association was
observed between 7-day cumulative averaged O3 concentrations and tHcy (percent
change for fasting tHcy 4.16 [95% CI: -1.76, 10.42] and percent change for
postmethionine-load tHcy -0.65 [95% CI: -5.66, 4.71] per 20 ppb increase in
24-h avg O3 concentrations). No evidence of effect modification by smoking was
observed.
Blood lipids and glucose metabolism markers
Chuang et al. (2010) conducted a population-based cross-sectional analysis of data
collected on 7,778 participants during the Taiwanese Survey on Prevalence of
Hyperglycemia, Hyperlipidemia, and Hypertension in 2001. Apolipoprotein B
(ApoB), the primary apolipoprotein among low-density lipoproteins, was associated
with 3-day avg O3 concentration at the p <0.10 level. The 5-day mean O3
concentration was associated with an increase in triglycerides at p <0.10. In addition,
the 1-, 3-, and 5-day mean O3 concentrations were associated with increased HbAlc
6-182
-------
levels (a marker used to monitor the degree of control of glucose metabolism) at the
p <0.05 level. The 5-day mean O3 concentration was associated with increased
fasting glucose levels (p <0.10). No association was observed between O3
concentration and ApoAl.
6.3.2.5 Myocardial Infarction (Ml)
The 2006 O3 AQCD did not report consistent results indicating an association
between short-term O3 exposure and MI. One study reported a positive association
between current day O3 concentration and acute MI, especially among the oldest age
group (55 to 64 year-olds) (Ruidavets et al., 2005b). No association was observed in
a case-crossover study of O3 concentration during the surrounding hours and MI
(Peters et al., 2001). Since the 2006 O3 AQCD, a few recent epidemiologic studies
have examined the association between O3 concentration and MI (Henrotin et al.,
2010: Richetal..201Q). arterial stiffness (Wu et al.. 2010) and ST-segment
depression (Delfino et al., 2011).
One of the studies conducted in the U.S. examined hospital admissions for first MI
and reported no association with O3 concentration (Rich et al.. 2010). More details
on this study are reported in the section on hospital admissions (Section 6.3.2.7).
A study performed in Dijon, France examined the association between O3
concentration and incident and recurrent MI (Henrotin et al.. 2010). The mean 8-hour
O3 concentration was 19.1 ppb (SD 12.2 ppb). Odds ratios for the association
between cumulative O3 concentrations and recurrent Mis were elevated but none of
the results were statistically significant (OR: 1.71 [95% CI: 0.91, 3.20] per 20 ppb
increase in 24-h avg O3 concentration for a cumulative lag of 1-3 days).
No association was observed for incident Mis. In analyses stratified by vascular risk
factors, positive associations were observed between 1-day lagged O3 concentration
and Mis (incident and recurrent combined) among those who reported having
hypercholesterolaemia (OR: 1.52 [95% CI: 1.08, 2.15] per 20 ppb increase in
24-h avg O3 concentration) and a slight inverse association was observed among
those who reported not having hypercholesterolaemia (OR: 0.69 [95% CI: 0.50, 0.94]
per 20 ppb increase in 24-h avg O3 concentration). In other stratified analyses
combining different vascular factors, only those containing individuals with
hypercholesterolaemia demonstrated a positive association; none were inverse
associations.
Wu et al. (2010) examined mail carriers aged 25-46 years and measured exposure to
O3 concentrations through personal monitors [mean O3 24.9 (SD 14.0) ppb]. Ozone
concentration was positively associated with arterial stiffness (percent change
11.24% [95% CI: 3.67, 19.62] per 40 ppb O3) and was robust to adjustment for
ultrafine PM.
A study performed in the Los Angeles basin reported on the association between O3
concentration and ST-segment depression, a measure representing cardiac ischemia
(Delfino et al., 2011). Study participants were nonsmokers, at least 65 years old, had
6-183
-------
a history of coronary artery disease, and were living in a retirement community.
Study periods included five consecutive days in both July to mid-October and mid-
October to February. Mean 24-hour O3 concentrations were 27.1 ppb (SD 11.5 ppb).
No association was observed between O3 concentration and ST-segment depression
of at least 1.0 mm during any of the exposure periods (i.e., 1-hour, 8-hours, 1-day,
2-day avg, 3-day avg, 4-day avg).
6.3.2.6 Blood Pressure
In the 2006 O3 AQCD, no epidemiologic studies examined O3-related effects on
blood pressure (BP). Recent studies have been conducted to evaluate this relationship
and overall the findings are inconsistent. The O3 concentrations for these studies are
listed in Table 6-33.
Table 6-33 Characterization of O3 concentrations (in ppb) from studies of
blood pressure.
Study*
Zanobetti et al. (2004)
Delfinoetal. (201 Ob)
Choi et al. (2007)
Chuanq et al. (2010)
Location
Boston,
Massachusetts
Los Angeles,
California
Incheon,
South Korea
Taiwan
Averaging Time
1-h
5-days
24-h
8-h
(warm season)
8-h
(cold season)
Mean Concentration
(Standard Deviation)
20
24
27.1 (11.5)
26.6(11.8)
17.5(7.3)
26.83 (9.7)
Upper Range of
Concentration
Max: 60.7
75th: 34.8
Max: 62.4
75th: 22.9
Max: 33.9
Max: 62.1
*Note: Studies presented in order of first appearance in the text of this section.
Zanobetti et al. (2004) examined the relationship between air pollutants and BP from
May 1999 to January 2001 for 631 repeat visits among 62 Boston, MA, residents
with CVD. In single-pollutant models, higher resting diastolic blood pressure (DBP)
was associated with the 5-day (0-4 days) averages of O3 concentration (RR: 1.03
[95% CI: 1.00, 1.05] per 20 ppb increase in 24-hour O3 concentrations). However,
this effect was no longer apparent when PM2.s was included in the model (data were
not presented) (Zanobetti et al.. 2004). Delfino et al. (201 Ob) examined 64 subjects
65 years and older with coronary artery disease, no tobacco smoke exposure, and
living in retirement communities in the Los Angeles air basin with hourly (up to 14-
hours/day) ambulatory BP monitoring for 5 days during a warm period (July-mid-
October) and 5 days during a cool period (mid-October-February). Investigators
assessed lags of 1, 4, and 8 hours, 1 day, and up to 9 days before each BP measure;
no evidence of an association was observed for O3 (change in BP associated with a
6-184
-------
20 ppb increase in 24-h avg O3 concentration was 0.67 [95% CI: -1.16, 2.51 for
systolic BP [SBP] and -0.25 [95% CI: -1.25, 0.75] for DBF) (Delfino et al.. 2010V).
Choi et al. (2007) conducted a cross-sectional study to investigate the relationship
between air pollutants and BP among 10,459 participants of the Inha University
Hospital health examination from 2001 to 2003. These individuals had no medical
history of cardiovascular disease or hypertension. Ozone concentration was
associated with an increase in SBP for 1-day lag in the warm season and similar
effect estimates were observed during the cold season but were not statistically
significant (quantitative results not provided). Associations between O3
concentration and DBP were present in the cold season but not the warm season
(quantitative results not provided). Chuang et al. (2010) conducted a similar type of
study among 7,578 participants of the Taiwanese Survey on Prevalence of
Hyperglycemia, Hyperlipidemia, and Hypertension in 2001. Investigators examined
1-, 3-, and 5-day avg O3 concentrations. An increase in DBP was associated with the
3-day mean O3 concentration (change in BP for a 20 ppb increase in 24-h avg O3
concentration was 0.61 [95% CI: 0.07, 1.14]) (Chuang et al.. 2010). Associations
were not observed for other days or with SBP.
6.3.2.7 Hospital Admissions and Emergency Department Visits
Upon evaluating the collective evidence for O3-related cardiovascular hospital
admissions and emergency department (ED) visits, the 2006 O3 AQCD concluded
that "a few studies observed positive O3 associations, largely in the warm season.
Overall, however, the currently available evidence is inconclusive regarding any
association between ambient O3 exposure on cardiovascular hospitalizations" (U.S.
EPA. 2006b). Table 6-34 provides information on the O3 concentrations reported in
each of the recent hospital admission and ED visit studies evaluated.
6-185
-------
Table 6-34 Characterization of O3 concentrations (in ppb) from studies of
hospital admissions and ED visits.
Study3
Peel et al. (2007)
Tolbert et al. (2007)
Katsouyanni et al. (2009)
Richetal. (2010)
Cakmak et al. (2006a)
Stieb et al. (2009)
Szvszkowicz (2008)
Villeneuve et al. (2006a)
Svmons et al. (2006)
Welleniusetal. (2005)
Zanobetti and Schwartz (2006)
Yang (2008)
Lee et al. (2007)
Chan et al. (2006)
Chiu and Yang (2009)
Lee et al. (2008a)
Wong et al. (2009)
Bell et al. (2008)
Buadong et al. (2009)
Lee et al. (2003b)
Azevedoetal. (2011)
Linares and Diaz (2010)
Middleton et al. (2008)
Location
Atlanta, GA
Atlanta, GA
12 Canadian cities
8 European cities
14 United
Statescities
New Jersey
10 Canadian cities
7 Canadian cities
Edmonton, Canada
Edmonton, Canada
Baltimore, MD
Allegheny County,
PA
Boston, MA
Taipei, Taiwan
Kaohsiung, Taiwan
Taipei, Taiwan
Taipei, Taiwan
Taipei, Taiwan
Hong Kong
Taipei, Taiwan
Bangkok, Thailand
Seoul, Korea
Portugal
Madrid, Spain
Nicosia, Cyprus
Averaging
Time
8-h max
warm season
8-h max
warm season
1-h
1-h
1-h
24-h
1-h max
24-h
24-h
24-h
24-h
warm season
24-h
cold season
8-h
warm season
24-h
24-h
24-h
24-h
1-h max
24-h
24-h
8-h
24-h
1-h
1-h max
1-h
24-h
8-h max
Mean Concentration
(Standard Deviation)
55.6 (23.8)
53.0
6.7-8.3*
11.0-38.1*
34.9-60.0*
NR
17.4
18.4
18.6(9.3)
17(9.1)
21 .8 (8)
12.2(7.4)
31.0(20.0)
24.3(12.2)
22.4*
21.0
26.5
50.9 (26.4)
23.0
21.0
18.5(11.5)
21.4
14.4(3.2)
36.0(18.6)
NR
17.4(8.9)
28.7 - 54.9
Upper Range of
Concentration
75th: 67.0
Max: 147.5
75th: 8.4-1 2.4
75th: 15.3-49.4
75th: 46.8-68.8
75th: 23.5
75th: 27.0
75th: 17.0
Max: 120.0
75th: 32.0
75th: 31 .0
75th: 26.3
Max: 62.8
75th: 35.5
Max: 83.0
Max: 150.3
75th: 28.7
Max: 62.8
75th: 26.4
Max: 62.8
75th: 25.4
Max: 48.3
Max: 53.4
Max: 41 .9
75th: 44.9
6-186
-------
Study3
Turner et al. (2007)
Ballesteretal. (2006)
De Pablo et al. (2006)
Von Klot et al. (2005)
Oudinetal. (2010)
Halonen et al. (2009)
Larrieu et al. (2007)
Barnett et al. (2006)
Hinwood et al. (2006)
Lanki et al. (2006)
Hosseinpoor et al. (2005)
Simpson et al. (2005)
Dennekamp et al. (2010)
Silvermanetal. (2010)
Location
Sydney, Australia
14 Spanish cities
Castilla-Leon, Spain
5 European cities
Scania, Sweden
Helsinki, Finland
8 French cities
4 Australian cities
Perth, Australia
5 European cities
Tehran, Iran
4 Australian cities
Melbourne, Australia
New York City, NY
Averaging
Time
24-h
8-h
warm season
24-h
8 h max
warm season
24-h
8-h max
warm season
8-h max
warm season
8-h
8-h max
8-h max
warm season
8-h max
1-h max
24-h
8-h max
Mean Concentration
(Standard Deviation)
28
24.2 - 44.3
23.2-33.6
16.4-28.0
30.5
35.7*
34.2-53.1
19.0-28.5
25.9 (6.5)
31 .7 - 57.2*
4.9 (4.8)
24.4-33.8
13.34
28*
Upper Range of
Concentration
75th: 33
75th: 42.1
Max: 79.6
Max: 58.4-86.8
75th: 7.2
Max: 99.0
Max: 96.0-111.5
75th: 16.93
75th: 40
"Notes: Median presented (information on mean not given); NR: Not reported;
text of this section.
Studies presented in order of first appearance in the
Multiple recent studies of O3 concentration and cardiovascular hospital admissions
and ED visits have been conducted in the U.S. and Canada. Peel et al. (2007) used a
case-crossover framework (using a time-stratified approach matching on day of the
week in the calendar month of the event) to assess the relationship between air
pollutants and cardiovascular disease ED visits among those with and without
secondary comorbid conditions (hypertension, diabetes, chronic obstructive
pulmonary disease [COPD], congestive heart failure [CHF], and dysrhythmia). Data
on over 4 million ED visits from 31 hospitals were collected from January 1993 to
August 2000. Ozone was monitored from March to October. This study was a re-
analysis of a time series study conducted to assess the main effects of air pollutants
on cardiovascular ED visits in Atlanta, GA (Tolbert et al.. 2007: Metzger et al..
2004). In the initial study, no evidence of associations was observed between O3
concentration and all CVD visits or visits for CVD subgroups, such as dysrhythmia,
CHF, ischemic heart disease (IHD), and peripheral vascular and cerebrovascular
disease. The relative risk for all CVD visits was 1.01 (95% CI: 0.98, 1.04) for a
30 ppb increase in the 3-day moving avg (lags 0-2 days) of 8-hour O3 concentration
(Metzger et al.. 2004). Similar to the initial investigation using a time-series analysis,
no evidence of an association was observed between short-term O3 exposure and
CVD visits at lag 0-2 among the entire population using the case-crossover design
(Peel et al.. 2007). However, the relationship between O3 concentration and
6-187
-------
peripheral and cerebrovascular disease visits was stronger among patients with
comorbid COPD (OR: 1.29 [95% CI: 1.05-1.59] per 30 ppb, lag 0-2 days) as
compared to patients without COPD (OR: 1.01 [95% CI: 0.96-1.06] per 30 ppb, lag
0-2 days). The same research group expanded upon the number of Atlanta hospitals
providing ED visit data (41 hospitals) as well as the length of the study period (1993-
2004) (Tolbert et al.. 2007). Again, models assessing the health effects of O3
concentration utilized data collected from March through October. Similar to the
results presented by Metzger et al. (2004) and Peel et al. (2007) among the entire
study population, no evidence of associations was observed for O3 concentration and
CVD visits (Tolbert et al.. 2007).
Existing multicity studies in North America and Europe were evaluated under a
common framework in the Air Pollution and Health: A European and North
American Approach (APHENA) study (Katsouyanni et al., 2009). One component of
the study examined the relationship between short-term O3 exposure and CVD
hospital admissions among individuals 65 years of age and older. The study
presented multiple models but this section focuses on the results for the models that
used 8 df to account for temporal trends and natural splines (see Section 6.2.7.2 for
additional explanation). Across the study locations, no associations were observed
between O3 concentration and CVD hospital admissions at lags 0-1, lag 1, or a
distributed lag of 0-2. Additionally, there was no evidence of an association when
restricting the analysis to the summer months.
A study of hospital admissions for MI was performed using a statewide registry from
New Jersey between January 2004 and December 2006 (Rich et al.. 2010). Using a
case-crossover design, the association between the previous 24-hours O3
concentration and transmural infarction (n = 1,003) was examined. No association
was observed (OR: 0.94 [95% CI: 0.79, 1.13] per 20 ppb increase in 24-h avg O3
concentration) and this did not change with the inclusion of PM2.s in the model.
Cakmak et al. (2006a) investigated the relationship between gaseous air pollutants
and cardiac hospitalizations in 10 large Canadian cities using a time-series approach.
A total of 316,234 hospital discharge records for primary diagnosis of congestive
heart failure, ischemic heart disease, or dysrhythmia were obtained from April 1993
through March 2000. Correlations between pollutants varied substantially across
cities, which could partially explain discrepancies in effect estimates observed across
the cities. In addition, pollutant lags differed across cities; the average lag for O3 was
2.9 days. The pooled effect estimate for a 20 ppb increase in the daily 1-h max O3
concentration and the percent change in hospitalizations among all 10 cities was 2.3
(95% CI: 0.11, 4.50) in an all-year analysis. The authors reported no evidence of
effect modification by sex, neighborhood-level education, or neighborhood-level
income. A similar multicity time-series study was conducted using nearly 400,000
ED visits to 14 hospitals in seven Canadian cities from 1992 to 2003 (Stieb et al..
2009). Primary analyses considered daily O3 single day lags of 0-2 days; in addition,
sub-daily lags of 3-h avg concentrations up to 12 hours before presentation to the ED
were considered. Seasonal variation was assessed by stratifying analyses by warm
and cold seasons. No evidence of associations between short-term O3 exposure and
6-188
-------
CVD ED visits was observed. One negative, statistically significant association was
reported between a 1-day lag of O3 concentration and visits for angina/myocardial
infarction. Ozone concentration was negatively correlated with many of the other
pollutants, particularly during the cold season.
The effect of air pollution on daily ED visits for ischemic stroke (n = 10,881 visits)
in Edmonton, Canada was assessed from April 1992 through March 2002
(Szyszkowicz. 2008). A 26.4% (95% CI: 3.16-54.5) increase in stroke ED visits was
associated with a 20 ppb increase in 24-hour average O3 concentration at lag 1
among men aged 20-64 years in the warm season. No associations were present
among women or among men age 65 and older. In addition, no associations were
observed for the cold season or for other lags (lag 0 or lag 2). A similar investigation
over the same time period in Edmonton, Canada, assessed the relationship between
air pollutants and ED visits for stroke (ischemic stroke, hemorrhagic stroke, and
transient ischemic attack) among those 65 years of age and older using a case-
crossover framework (Villeneuve et al., 2006a). No evidence of association was
reported for O3 concentration and stroke hospitalization in single or copollutant
models (Villeneuve et al., 2006a).
Additional studies in the U.S. reported no evidence of an association between O3
concentrations and ED visits, hospitalizations, or symptoms leading to
hospitalization (Symons et al.. 2006: Zanobetti and Schwartz. 2006: Wellenius et al..
2005). Symons et al. (2006) used a case-crossover framework to assess the
relationship between air pollutants and the onset of symptoms (dyspnea) severe
enough to lead to hospitalization (through the ED) for congestive heart failure.
The study was conducted from April to December of 2002 in Baltimore, Maryland.
Exposures were assigned using 3 index times: 8-hour and 24-hour periods prior to
symptom onset and date of hospital admission. No evidence of association was
reported for O3 concentrations. Although seasonal variation was not assessed, the
time frame for the study did not involve an entire year (April to December).
Wellenius et al. (2005) investigated the association between air pollutants and
congestive heart failure hospitalization among Medicare beneficiaries in Pittsburgh,
Pennsylvania from 1987 to 1999 utilizing a case-crossover framework. A total of
55,019 admissions from the emergency room with a primary discharge diagnosis of
CHF were collected. No evidence of an association was reported for O3
concentration and CHF hospitalization (Wellenius et al., 2005). Finally, Zanobetti
and Schwartz (2006) assessed the relationship between air pollutants and hospital
admissions through the ED for MI and pneumonia among patients aged 65 and older
residing in the greater Boston, MA, area (1995-1999) using a case-crossover
framework with control days in the same month matched on temperature. Pollution
exposures were assigned for the same day and for the mean of the exposure the day
of and the day before the admission. Ozone concentration was not associated with MI
admissions in all-year and seasonal analyses.
Several recent studies have examined the relationship between air pollution and CVD
hospital admissions and/or emergency department visits in Asia. Of note, some areas
of Asia have a more tropical climate than the U.S. and do not experience similar
6-189
-------
seasonal changes. In Taiwan, fairly consistent positive associations have been
reported for O3 concentration and congestive heart failure hospital admissions (for
single- and copollutant models) in Taipei on warm days (Yang, 2008) and in
Kaohsiung (Lee et al.. 2007); cerebrovascular disease ED visits (for lag 0 single- and
two-pollutant models but not other lags) in Taipei (Chan et al., 2006); and arrhythmia
ED visits in Taipei among those without comorbid conditions (Chiu et al.. 2009; Lee
et al.. 2008a) and in Taipei on warm days among those with and without comorbid
conditions (Lee et al.. 2008a). However, one study in Taiwan did not show an
association. Bell et al. (2008) reported no evidence of an association between O3
concentration and hospital admissions for ischemic heart disease or cerebrovascular
disease. Studies based in Asia but outside Taiwan were also performed. A Hong
Kong-based investigation (Wong et al.. 2009) reported no consistent evidence of a
modifying effect of influenza on the relationship between O3 concentration and CVD
admissions. Among elderly populations in Thailand, O3 concentration was associated
with CVD visits, but this association was not detected among younger age groups
(15-64) (Buadong et al.. 2009). Also, a study performed in Seoul, Korea reported a
positive association between O3 concentration and hospital admissions for ischemic
heart disease; the association was slightly greater among those over 64 years of age
(Lee et al.. 2003b).
Positive associations between short-term O3 exposure and CVD hospital admissions
and/or ED visits have been reported in other areas of the world as well (Azevedo et
al.. 2011: Linares and Diaz. 2010; Middleton et al.. 2008; Turner et al.. 2007;
Ballester et al.. 2006; De Pablo et al.. 2006; Von Klot et al.. 2005). although not
consistently; some studies reported no association (Oudin et al.. 2010; Halonen et al..
2009; Larrieu et al.. 2007; Barnett et al.. 2006; Hinwood et al.. 2006; Lanki et al..
2006; Hosseinpoor et al.. 2005; Simpson et al.. 2005).
A couple of studies (U.S. and Australia) have examined cardiac arrests where
emergency services attempted treatment/resuscitation. No evidence of an association
between O3 concentration and out-of-hospital cardiac arrest was observed
(Dennekamp et al.. 2010; Silverman et al.. 2010).
An increasing number of air pollution studies have investigated the relationship
between O3 concentrations and CVD hospital admissions and/or ED visits.
As summarized in the 2006 O3 AQCD, some, especially those reporting results
stratified by season (or temperature) or comorbid conditions have reported positive
associations. However, even studies performing these stratified analyses are not
consistent and the overall evidence remains inconclusive regarding the association
between short-term O3 exposure and CVD hospital admissions and ED visits.
The Hospital Admission (HA) and ED visit studies evaluated in this section are
summarized in Figure 6-22 through Figure 6-26. which depict the associations for
studies in which quantitative data were presented. Table 6-35 through Table 6-39
provide the numerical results displayed in the figures.
6-190
-------
R eference
Buadong et al. (2009)
Katsouyanni et al. (2009)
Katsouyanni et al. (2009)
Katsouyanni et al. (2009)
Middleton et al.(2008)
Fungetal. (2005)
Ballesteret al. (2001)
Petroeschevsky et al. (2001)
Linnet al. (2000)
Atkinson et al.(1999)
Wongetal.(1999a)
Wongetal.(1999b)
Prescottetal. (1998)
Poloniecki et al.(1997)
Halonen et al. (2009)
Katsouyanni et al. (2009)
Katsouyanni et al. (2009)
Katsouyanni et al. (2009)
Larrieuet al. (2007)
Peelet al. (2007)
Ballesteret al. (2006)
Chang et al.(2005)
Yang et al.(2004)
Wong etal. (1999b)
Chang et al.(2005)
Yang et al.(2004)
Wong etal. (1999a)
Wong etal. (1999b)
Cakmaketal. (2006)
Ballesteret al. (2001)
Morgan etal. (1998)
Larrieuet al. (2007)
Ballesteret al. (2006)
von Klot et al. (2005)
Bell etal. (2008)
Chanel al. (2006)
Ballesteret al. (2001)
Wong etal. (1999a)
Wongetal.(1999b)
Poloniecki etal. (1997)
Peelet al. (2007)
Wongetal.(1999b)
Wongetal.(1999b)
Location
Bangkok, Thailand
14 U.S. cities
12 Canadian cities
8 European cities
Nicosia, Cyprus
Windsor, Canada
Valencia, Spain
Brisbane, Australia
Los Angeles, CA
London, England
Hong Kong
Hong Kong
Edinburgh, Scotland
London, England
Helsinki, Finland
14 U.S. cities
12 Canadian cities
8 European cities
8 French cities
Atlanta, GA
14 Spanish cities
Taipei, Taiwan
Kaohsiung, Taiwan
Hong Kong
Taipei, Taiwan
Kaohsiung, Taiwan
Hong Kong
Hong Kong
10 Canadian cities
Valencia, Spain
Sydney, Australia
8 French cities
14 Spanish cities
5 European cities
Taipei, Taiwan
Taipei, Taiwan
Valencia, Spain
Hong Kong
Hong Kong
London, England
Atlanta, GA
Hong Kong
Hong Kong
Cardio vascular disease
Cardiac disease
Cerebrovascular disease
0.70 0.80 0.90 1.00 1.10
Effect Estimate
1.20
1.30 1.40 1.50
Note: Change in O3 standardized to 20 ppb for 24-h avg period, 30 ppb for 8-h avg period, and 40 ppb for 1 -h avg period (see
Section 2.5). Ozone concentrations in ppb. Seasons depicted by colors - black: all year; red: warm season; light blue: cold
season. Age groups of study populations were not specified or were adults with the exception of Katsouvanni et al. (2009). Fung
et al. (2005). Wong et al. (1999b). and Prescott et al. (1998). which included only individuals aged 65+. Studies organized by
outcome and season and then listed in descending order of publication date.
Figure 6-22 Effect estimate (95% Cl) per increment ppb increase in O3 for over
all cardiovascular ED visits or hospital admissions.
6-191
-------
Table 6-35 Effect estimate (95% Cl) per increment ppb increase in O3 for
overall cardiovascular ED visits or hospital admissions in studies
presented in Figure 6-22.
Study*
Atkinson et al. (1999)
Ballesteretal. (2006)
Ballesteretal. (2006)
Bell et al. (2008)
Buadong et al. (2009)
Cakmak et al. (2006a)
Chan et al. (2006)
Chang et al. (2005)
Fung et al. (2005)
Halonen et al. (2009)
Katsouvanni et al.
(2009)
Larrieu et al. (2007)
Linn et al. (2000)
Middleton et al. (2008)
Morgan et al. (1998)
Peel et al. (2007)
Petroeschevsky et al.
(2001 )
Polonieckietal. (1997)
Location
London, England
14 Spanish cities
Valencia, Spain
Taipei, Taiwan
Bangkok, Thailand
10 Canadian cities
Taipei, Taiwan
Taipei, Taiwan
Windsor, Canada
Helsinki, Finland
14 U.S. cities
12 Canadian cities
8 European cities
8 French cities
Los Angeles,
California
Nicosia, Cyprus
Sydney, Australia
Atlanta, GA
Brisbane,
Australia
London, England
Outcome
Cardiovascular disease
Cardiovascular disease
Cardiac disease
Cardiovascular disease
Cardiac disease
Cerebrovascular
disease
Cerebrovascular
disease
Cardiovascular disease
Cardiac disease
Cerebrovascular
disease
Cardiovascular disease
Cardiovascular disease
Cardiovascular disease
Cardiovascular disease
Cardiovascular disease
Cardiac disease
Cardiovascular disease
Cardiovascular disease
Cardiac disease
Cardiovascular disease
Cerebrovascular
disease
Cardiovascular disease
Cardiovascular disease
Cerebrovascular
disease
Averaging Time Effect Estimate (95% Cl)
8-h
8-h warm season
8-h warm season
8-h
8-h
8-h
24-h
1-h
1-h max
1-h max
24-h warm season
24-h cold season
1-h
8-h max warm
season
1 -h max
1 -h max warm
season
1-h max
1 -h max warm
season
1-h max
1 -h max warm
season
8-h max warm
season
24-h
8-h max
1-h max
8-h warm season
8-h warm season
8-h
8-h
8-h
1.03(1.00, 1.05)
1.04(1.02, 1.06)
1.04(1.01, 1.07)
0.94 (0.84, 1 .06)
0.88 (0.75, 1 .03)
0.86 (0.72, 1 .04)
0.94 (0.87, 1 .02)
1.01 (1.00, 1.02)
1 .02 (1 .00, 1 .04)
1.02(1.01, 1.03)
1 .42 (1 .33, 1 .50)
1.15(1.04, 1.27)
1.02(0.92, 1.13)
1.05(0.96, 1.14)
1.01 (0.99, 1.03)
1 .00 (0.97, 1 .03)
1 .00 (0.95, 1 .04)
0.98 (0.94, 1 .02)
0.99 (0.96, 1 .02)
0.98 (0.94, 1 .03)
1.01 (0.98, 1.04)
0.99 (0.98, 1 .00)
1.09(1.00, 1.18)
1 .02 (0.99, 1 .05)
1 .00 (0.98, 1 .02)
1 .03 (0.97, 1 .08)
0.96(0.92, 1.01)
0.97(0.93, 1.01)
0.98 (0.95, 1 .02)
6-192
-------
Study*
Prescott et al. (1998)
Von Klot et al. (2005)
Wong etal. (1999b)
Wong etal. (1999a)
Yana et al. (2004)
Location Outcome
U±T Cardiovascular disease
5 European cities Cardiac disease
Hong Kong
Cerebrovascular
disease
Cardiovascular disease
Cerebrovascular
disease
Kaohsiung,
i aiwan
Averaging Time Effect Estimate (95% Cl)
24-h
8-h max warm
season
24-h
24-h cold season
24-h
24-h
24-h warm season
24-h cold season
24-h
24-h warm season
24-h cold season
24-h warm season
24-h cold season
0.89 (0.78,
1.11 (1.00,
1.08(1.03,
1.15(1.04,
0.95 (0.90,
1.02(1.03,
1.01 (0.96,
1.06(1.02,
0.99 (0.95,
0.98 (0.90,
1.02(0.96,
1.33(1.26,
1.05(0.96,
1.00)
1.22)
1.13)
1.26)
1.01)
1.06)
1.06)
1.11)
1.04)
1.08)
1.10)
1.40)
1.15)
'Studies included in Figure 6-22.
Note: Change in O3 standardized to 20 ppb for 24-hour averaging period, 30 ppb for 8-hour averaging period, and 40 ppb for 1 -hour
averaging period (see Section 2.5). Ozone concentrations in ppb. Age groups of study populations were not specified or were
adults with the exception of Katsouvanni et al. (2009). Fung et al. (2005). Wong etal. (1999a). and Prescott et al. (1998). which
included only individuals aged 65+. Studies listed in alphabetical order.
Warm season defined as: March-October (Peel et al.. 2007). May-October (Ballester et al.. 2005: Wong etal.. 1999a). May-
September (Halonen et al.. 2009). April-September (Katsouvanni etal.. 2009: Larrieu et al.. 2007: Von Klot et al.. 2005).
> 20°C (Chang et al.. 2005) and > 25°C (Yang et al.. 2004). Cold season defined as: November-April (Wong etal.. 1999a).
<20°C (Chang et al.. 2005) and <25°C (Yang et al.. 2004). December-March (Wong etal.. 1999b)
6-193
-------
Reference
Stiebetal.(2009)
Welleniusetal.(2005)
Wongetal.(1999a)
Wongetal.(1999b)
Polonieckietal.(1997)
Yang (2008)
Peel etal. (2007)
Lee etal. (2007)
Sy mo ns etal. (2006)
Wong etal. (1999b)
Yang (2008)
Lee etal. (2007)
Wong etal. (1999b)
Location
7Canadian cities
Allegheny county, PA
Hong Kong
Hong Kong
London, England
Taipei, Taiwan
Atlanta, GA
Kaohsiung, Taiwan
Baltimore, MD
Hong Kong
Taipei, Taiwan
Kaohsiung, Taiwan
Hong Kong
0.40 0.60 0.80 1.00 1.20 1.40
Effect Estimate
1.60
1.80
2.00
Note: Change in O3 standardized to 20 ppb for 24-hour averaging period, 30 ppb for 8-hour averaging period, and 40 ppb for 1 -hour
averaging period (see Section 2.5). Ozone concentrations in ppb. Seasons depicted by colors: black: all year; red: warm season;
light blue: cold season. Outcomes were all congestive heart failure, with the exception of Svmons et al. (2006). which examined
onset of congestive heart failure symptoms leading to a heart attack. Age groups of study populations were not specified or were
adults with the exception of Wellenius et al. (2005) and Wong etal. (1999a). which included only individuals aged 65+. Studies
organized by outcome and season and then listed in descending order of publication date.
Figure 6-23 Effect estimate (95% Cl) per increment ppb increase in O3 for
congestive heart failure ED visits or hospital admissions.
6-194
-------
Table 6-36 Effect estimate (95% Cl) per increment ppb increase in O3 for
congestive heart failure ED visits or hospital admissions for
studies in Figure 6-23.
Study*
Lee et al. (2007)
Peel et al. (2007)
Poloniecki et al.
(1997)
Stieb et al. (2009)
Svmons et al.
(2006)
Welleniuset al.
(2005)
Wong et al.
(1999a)
Yang (2008)
Wong et al.
(1999b)
Location
Kaohsiung, Taiwan
Atlanta, GA
London, England
7 Canadian cities
Baltimore, MD
Allegheny county, PA
Hong Kong
Taipei, Taiwan
Hong Kong
Outcome
Congestive heart failure
Congestive heart failure
Congestive heart failure
Congestive heart failure
Congestive heart failure
Onset of congestive heart
failure symptoms leading to
heart attack
Congestive heart failure
Congestive heart failure
Congestive heart failure
Congestive heart failure
Congestive heart failure
Averaging Time
24-h warm season
24-h cold season
8-h warm season
8-h
24-h
8-h warm season
24-h
24-h
24-h warm season
24-hcold season
24-h warm season
24-h cold season
24-h
Effect Estimate
(95% Cl)
1.25(1.15,1.36)
1.24(1.09,1.41)
0.94 (0.89, 1 .00)
0.99 (0.95, 1 .03)
1 .03 (0.98, 1 .07)
0.83(0.49,1.41)
0.98(0.96, 1.01)
1.11 (1.04,1.80)
1.09(0.96, 1.23)
1.16(1.06,1.27)
1.39(1.27,1.51)
0.61 (0.52, 0.73)
1.25(1.11,1.41)
'Studies include those from Figure 6-23.
Note: Change in O3 standardized to 20 ppb for 24-hour averaging period, 30 ppb for 8-hour averaging period, and 40 ppb for
1-hour averaging period (see Section 2.5). Ozone concentrations in ppb. Outcomes were all congestive heart failure, with
the exception of Svmons et al. (2006). which examined onset of congestive heart failure symptoms leading to a heart attack.
Age groups of study populations were not specified or were adults with the exception of Welleniuset al. (2005) and Wong et
al. (1999a). which included only individuals aged 65+. Studies listed in alphabetical order.
Warm season defined as: March-October (Peel et al., 2007), April-November (Svmons et al., 2006), May-October (Wong et
al.. 1999a) > 20°C (Yang. 2008). and >25°C (Lee etal.. 2007). Cold season defined as: November-April (Wong et al..
1999a). <20°C (Yang. 2008). and <25°C (Lee et al.. 2007).
6-195
-------
Reference
Buadong etal. (2009)
Belletal.(200B)
Lee etal. (2003)
Atkinson etal. (1999)
Wongetal.[1999a)
Wongetal.[1999bj
Larrieu etal. (2007)
Peel etal. (2007)
Lee etal. (2003)
Wongetal.[1999bj
Wongetal.(1999b)
Halonen etal. (2009)
Rich etal. (2010)
Buadong etal. (2009)
Stiebetal.(2009)
Zanobetti and Schwartz (2006)
Polonieckietal.(1997)
Lanki etal. (2006)
von Klot etal. (2005)
Hosseinpoor etal. (2005)
Poloniecki etal. (1997)
von Klot etal. (2005)
Location
Bangkok. Thailand H
Seoul. Korea
London. England — • —
HongKong —
S French cities —
Atlanta. GA — 4
Seoul. Korea
Bangkok. Thailand — • —
/Canadian cities — <
London. England — •-
5 European cities — • —
London. England — •-
5 European cities
Ischemia heart disease
H
-»
-•
>
Coronafy h&art disease
Myocardiaf infarction
>
Angina pectoris
0.5 0.7 0.9 1.1 1.3 1.5
Effect Estimate
Note: Change in O3 standardized to 20 ppb for 24-hour averaging period, 30 ppb for 8-hour averaging period, and 40 ppb for 1 -hour
averaging period (see Section 2.5). Ozone concentrations in ppb. Seasons depicted by colors: black: all year; red: warm season;
light blue: cold season. Age groups of study populations were not specified or were adults with the exception of Wong et al.
(1999a) and Atkinson et al. (1999). which included only individuals aged 65+. Studies organized by outcome and season and then
listed in descending order of publication date.
Figure 6-24 Effect estimate (95% Cl) per increment ppb increase in O$ for
ischemic heart disease, coronary heart disease, myocardial
infarction, and angina pectoris ED visits or hospital admissions.
6-196
-------
Table 6-37 Effect estimate (95% Cl) per increment ppb increase in O3 for
ischemic heart disease, coronary heart disease, myocardial
infarction, and angina pectoris ED visits or hospital admissions for
studies presented in Figure 6-24.
Study*
Atkinson etal. (1999)
Bell et al. (2008)
Buadong et al. (2009)
Halonen et al. (2009)
Hosseinpoor et al. (2005)
Lanki et al. (2006)
Larrieu et al. (2007)
Lee et al. (2003b)
Peel et al. (2007)
Polonieckietal. (1997)
Rich etal. (2010)
Stieb et al. (2009)
Von Klot et al. (2005)
Wong etal. (1999a)
Wong etal. (1999b)
Zanobetti and Schwartz
(2006)
Location
London, England
Taipei, Taiwan
Bangkok, Thailand
Helsinki, Finland
Tehran, Iran
5 European cities
8 French cities
Seoul, Korea
Atlanta, GA
London, England
New Jersey
7 Canadian cities
5 European cities
Hong Kong
Hong Kong
Boston, MA
Outcome
Ischemic heart disease
Ischemic heart disease
Ischemic heart disease
Myocardial infarction
Coronary heart disease
Angina
Myocardial infarction
Ischemic heart disease
Ischemic heart disease
Ischemic heart disease
Ischemic heart disease
Myocardial infarction
Angina
Myocardial infarction
Myocardial infarction
Myocardial infarction
Angina
Ischemic heart disease
Ischemic heart disease
Myocardial infarction
Averaging Time
8-h
24-h
1-h
1-h
8-h max
warm season
8-h max
8-h max
warm season
8-h max
warm season
1-h max
1-h max
warm season
8-h warm season
8-h
8-h
24-h
2-h
8-h max
warm season
8-h max
warm season
24-h
24-h warm season
24-h cold season
24-h
24-h
Effect Estimate
(95% Cl)
0.97(0.94, 1.01)
1.01 (0.91, 1.12)
1 .00 (0.98, 1 .02)
0.97(0.94, 1.01)
0.99 (0.79, 1 .25)
0.80 (0.70, 0.92)
0.96(0.92, 1.01)
1 .02 (0.98, 1 .07)
1.07(1.02, 1.13)
1.07(1.00, 1.17)
1 .00 (0.96, 1 .05)
0.98 (0.94, 1 .02)
0.98 (0.94, 1 .03)
0.94(0.79, 1.13)
1 .00 (0.96, 1 .04)
1.00(0.83, 1.21)
1.19(1.05, 1.35)
1.01 (0.94, 1.06)
1.02(0.94, 1.11)
1 .02 (0.95, 1 .09)
1 .03 (0.98, 1 .08)
0.98 (0.92, 1 .03)
'Studies included from Figure 6-24.
Note: Change in O3 standardized to 20 ppb for 24-hour averaging period, 30 ppb for 8-hour averaging period, and 40 ppb for 1 -hour
averaging period (see Section 2.5). Ozone concentrations in ppb. Age groups of study populations were not specified or were
adults with the exception of Wong et al. (1999a) and Atkinson et al. (1999). which included only individuals aged 65+. Studies
listed in alphabetical order.
Warm season defined as: March-October (Peel et al.. 2007). June-August (Lee et al.. 2003b). May-September (Halonen et al..
2009). May-October (Buadong etal.. 2009). and April-September (Larrieu et al.. 2007: Lanki etal.. 2006: Von Klot et al.. 2005).
Cold season defined as: November-April (Buadong et al.. 2009).
6-197
-------
Reference
Chanetal.(2006)
Halo nenetal. (2009)
Larrieuetal.(2007)
Chanetal.(2006)
Villeneuve et al. (2006)
Villeneuve et al. (2006)
Villeneuve et al. (2006)
Chanetal.(2006)
Villeneuve et al. (2006)
Villeneuve et al. (2006)
Villeneuve etal. (2006)
Villeneuve etal. (2006)
Villeneuve et al. (2006)
Villeneuve et al. (2006)
Location
All
Taipei, Taiwan
8 French cities •-
Taipei, Taiwan
^ m
i-
Ischemic
-•—
Hemorrhagic
Transient ischemic
0.5
0.7
0.9 1.1
Effect Estimate
1.3
1.5
Note: Change in O3 standardized to 20 ppb for 24-hour averaging period, 30 ppb for 8-hour averaging period, and 40 ppb for 1 -hour
averaging period (see Section 2.5). Ozone concentrations in ppb. Seasons depicted by colors: black: all year; red: warm season;
light blue: cold season. Age groups of study populations were not specified or were adults with the exception of Villeneuve etal.
(2006a). which included only individuals aged 65+, and.Chan et al. (2006). which included only individuals aged 50+. Studies
organized by outcome and season and then listed in descending order of publication date.
Figure 6-25 Effect estimate (95% Cl) per increment ppb increase in O$ for
stroke ED visits or hospital admissions.
6-198
-------
Table 6-38 Effect estimate (95% Cl) per increment ppb increase in O3 for
stroke ED visits or hospital admissions for studies presented in
Figure 6-25.
Study*
Chan et al. (2006)
Halonen et al. (2009)
Larrieu et al. (2007)
Villeneuve et al.
(2006a)
Location Outcome
All/non-specified stroke
Taipei, Taiwan Ischemic stroke
Hemorrhagic stroke
Helsinki, Finland All/non-specified stroke
8 French cities All/non-specified stroke
Ischemic stroke
CaLT' Hemorrhagic stroke
Transient ischemic
stroke
Averaging Time
1-h max
1-h max
1-h max
8-h max warm season
8-h max warm season
24-h
24-h warm season
24-h cold season
24-h
24-h warm season
24-h cold season
24-h
24-h warm season
24-h cold season
Effect Estimate
(95% Cl)
1.01 (0.99, 1
1 .03 (0.99, 1
0.99 (0.92, 1
1 .08 (0.83, 1
0.98 (0.93, 1
1 .00 (0.88, 1
1 .09 (0.91 , 1
0.98 (0.80, 1
1 .02 (0.87, 1
1.12(0.88, 1
0.97 (0.76, 1
0.98 (0.87, 1
0.85 (0.70, 1
1.11 (0.93, 1
.03)
.07)
.06)
.41)
.02)
.13)
.32)
.18)
.20)
.43)
.22)
.10)
.01)
.32)
'Studies included from Figure 6-25.
Note: Change in O3 standardized to 20 ppb for 24-hour averaging period, 30 ppb for 8-hour averaging period, and 40 ppb for 1 -hour
averaging period (see Section 2.5). Ozone concentrations in ppb. Age groups of study populations were not specified or were
adults with the exception of Villeneuve et al. (2006a). which included only individuals aged 65+, and Chan et al. (2006). which
included only individuals aged 50+. Studies listed in alphabetical order.
Warm season defined as: May-September (Halonen et al.. 2009). and April-September (Larrieu et al.. 2007: Villeneuve et al..
2006a). Cold season defined as: October-March (Villeneuve et al.. 2006a).
6-199
-------
Reference
Stiebetal.(2009)
Peel etal. (2007)
Location
/Canadian cities
Atlanta, GA
Buadongetal. (2009) Bangkok,Thailand
Wong etal. (1999b) Hong Kong
Poloniecki etal. (1997) London, England
Halonen etal. (2009) Helsinki, Finland
Wong etal. (1999b) Hong Kong
Wong etal. (1999b) Hong Kong
Dysrhythmia
Arrhythmia
0.70
0.80
0.90
1.00 1.10
Effect Estimate
1.20
1.30
1.40
Note: Change in O3 standardized to 20 ppb for 24-hour averaging period, 30 ppb for 8-hour averaging period, and 40 ppb for 1 -hour
averaging period (see Section 2.5). Ozone concentrations in ppb. Seasons depicted by colors: black: all year; red: warm season;
light blue: cold season. Age groups of study populations were not specified or were adults with the exception of Wong et al.
(1999a), which included only individuals aged 65+. Studies organized by outcome and season and then listed in descending order
of publication date.
Figure 6-26 Effect estimate (95% Cl) per increment ppb increase in O$ for
arrhythmia and dysrhythmia ED visits or hospital admissions.
6-200
-------
Table 6-39 Effect estimate (95% Cl) per increment ppb increase in O3 for
arrhythmia and dysrhythmia ED visits or hospital admissions for
studies presented in Figure 6-26.
Study
Buadong et al. (2009)
Halonen et al. (2009)
Peel et al. (2007)
Polonieckietal. (1997)
Stieb et al. (2009)
Wongetal. (1999a)
Location
Bangkok, Thailand
Helsinki, Finland
Atlanta, GA
London, England
7 Canadian cities
Hong Kong
Outcome
Arrhythmia
Arrhythmia
Dysrhythmia
Arrhythmia
Dysrhythmia
Arrhythmia
Averaging Time
1-h
8-h max warm season
8-h warm season
8-h
24-h
24-h
24-h warm season
24-h cold season
Effect Estimate
(95% Cl)
0.99 (0.95, 1
1 .04 (0.80, 1
1.01 (0.97, 1
1 .02 (0.96, 1
1 .02 (0.95, 1
1 .06 (0.99, 1
1.10(0.96, 1
1.11 (1.01, 1
.04)
.35)
.06)
.07)
.09)
.12)
.26)
.23)
'Studies included from Figure 6-26.
Note: Change in O3 standardized to 20 ppb for 24-hour averaging period, 30 ppb for 8-hour averaging period, and 40 ppb for 1 -hour
averaging period (see Section 2.5). Ozone concentrations in ppb. Age groups of study populations were not specified or were
adults with the exception of (Wong et al.. 1999a). which included only individuals aged 65+. Studies listed in alphabetical order.
Warm season defined as: March-October (Peel et al.. 2007). May-October (Wong et al.. 1999a) and May-September (Halonen et
al.. 2009). Cold season defined as: November-April (Wong et al.. 1999a).
6.3.2.8 Cardiovascular Mortality
As discussed within this section (Section 6.3). epidemiologic studies provide
inconsistent evidence of an association between short-term O3 exposure and
cardiovascular effects. However, toxicological studies have demonstrated
O3-induced cardiovascular effects, specifically enhanced atherosclerosis and altered
vascular function, which could lead to death. The 2006 O3 AQCD provided
evidence, primarily from single-city studies, of consistent positive associations
between short-term O3 exposure and cardiovascular mortality. Recent multicity
studies conducted in the U.S., Canada, and Europe further support the association
between short-term O3 exposure and cardiovascular mortality.
As discussed in Section 6.2.7.2. the APHENA study (Katsouvanni et al.. 2009) also
examined associations between short-term O3 exposure and mortality and found
consistent positive associations for cardiovascular mortality in all-year analyses.
However, in analyses restricted to the summer season, results were more variable
with no evidence of an association in the Canadian dataset in the population
<75 years of age, and evidence of associations persisting or increasing in magnitude
in the Canadian (population > 75 years of age), U.S., and European datasets.
Additional multicity studies from the U.S. (Zanobetti and Schwartz. 2008b). Europe
(Samoli et al.. 2009). Italy (Stafoggia et al.. 2010). and Asia (Wong et al.. 2010) that
conducted summer season and/or all-year analyses provide additional support for an
association between short-term O3 exposure and cardiovascular mortality
(Figure 6-27).
6-201
-------
Of the studies evaluated, only the APHENA study (Katsouyanni et al.. 2009) and the
Italian multicity study (Stafoggia et al., 2010) conducted an analysis of the potential
for copollutant confounding of the O3-cardiovascular mortality relationship. In the
European dataset, when focusing on the natural spline model with 8 df/year
(Section 6.2.7.2) and lag 1 results in order to compare results across study locations
(Section 6.6.2.1). cardiovascular mortality risk estimates were robust to the inclusion
of PMio in copollutant models in all-year analyses with more variability in the
Canadian and U.S. datasets (i.e., cardiovascular O3 mortality risk estimates were
reduced or increased in copollutant models). In summer season analyses,
cardiovascular O3 mortality risk estimates were robust in the European dataset and
attenuated but remained positive in the U.S. dataset. Similarly, in the Italian multicity
study (Stafoggia et al.. 2010). which was limited to the summer season,
cardiovascular mortality risk estimates were robust to the inclusion of PMi0 in
copollutant models. Based on the APHENA and Italian multicity results, O3
cardiovascular mortality risk estimates appear to be robust to inclusion of PMi0 in
copollutant models. However, in the U.S. and Canadian datasets there was evidence
that O3 cardiovascular mortality risk estimates are moderately to substantially
sensitive (e.g., increased or attenuated) to PMi0. The mostly every-6th-day sampling
schedule for PMi0 in the Canadian and U.S. datasets greatly reduced their sample
size and limits the interpretation of these results.
6.3.2.9 Summary of Epidemiologic Studies
Overall, the available body of evidence examining the relationship between short-
term exposures to O3 concentrations and cardiovascular morbidity is inconsistent.
Across studies, different definitions, i.e., I CD-9 diagostic codes were used for both
all-cause and cause-specific cardiovascular morbidity (Table 6-35. Table 6-36.
Table 6-37. Table 6-38. and Table 6-39). which may contribute to inconsistency in
results. However, within diagnostic categories, no consistent pattern of association
was found with O3. Generally, the studies summarized in this section used nearest air
monitors to assess O3 concentrations, with a few exceptions that used modeling or
personal exposure monitors (these exceptions were noted throughout the previous
sections). The inconsistencies in the associations observed between short-term O3
and CVD morbidities are unlikely to be explained by the different exposure
assignment methods used (see Section 4.6). The wide variety of biomarkers
considered and the lack of consistency among definitions used for specific
cardiovascular disease endpoints (e.g., arrhythmias, HRV) make comparisons across
studies difficult. Despite the inconsistent evidence for an association between O3
concentration and CVD morbidity, mortality studies indicate a consistent positive
association between short-term O3 exposure and cardiovascular mortality in multicity
studies and a multicontinent study.
6-202
-------
6.3.3 Toxicology
In the previous O3 AQCDs (U.S. EPA. 2006b. 1996a) experimental animal studies
have reported relatively few cardiovascular system alterations after exposure to O3
and other photochemical oxidants. The limited amount of research directed at
examining O3-induced cardiovascular effects has primarily found alterations in heart
rate (HR), heart rhythm, and BP after O3 exposure. Although O3 induced changes in
HR and core temperature (TCo) in a number of rat studies, these responses have not
been reported or extensively studied in humans exposed to O3 and may be unique to
rodents.
According to recent animal toxicology studies, short-term O3 exposure induces
vascular oxidative stress and proinflammatory mediators, alters HR and HRV, and
disrupts the regulation of the pulmonary endothelin system (study details are
provided in Table 6-40. A number of these effects were variable between strains
examined, suggesting a genetic component to development of O3 induced
cardiovascular effects. Further, recent studies provide evidence that extended O3
exposure enhances the risk of ischemia-reperfusion (I/R) injury and atherosclerotic
lesion development. Still, few studies have investigated the role of O3 reaction
products in these processes, but more evidence is provided for elevated inflammatory
and reduction-oxidation (redox) cascades known to initiate these cardiovascular
pathologies.
Heart Rate, Rhythm, and Heart Rate Variability
Studies (Arito etal.. 1992: Arito et al.. 1990: Uchiyama and Yokovama. 1989:
Yokoyama et al.. 1989: Uchiyama et al.. 1986) report O3 exposure (0.2-1.0 ppm, 3
hours to 3 days) in rats decreased TCo, HR, and mean arterial pressure (MAP).
In addition, O3 exposure (0.1-1.0 ppm, 3 hours to 3 days) in rats induced
arrhythmias, including increased PR interval and QRS complex, premature atrial
contraction, and incomplete A-V block (Arito et al.. 1990: Yokovama et al.. 1989:
Uchiyama et al.. 1986). The effects were more pronounced in adult and awake rats
than in younger or sleeping animals, whereas no sex-related differences were noted
in these O3 induced outcomes (Uchiyama et al.. 1986). However, these
cardiovascular responses to O3, including decreased TCo and HR, could be
attenuated by increased ambient temperatures and environmental stress and exhibited
adaptation (Watkinson et al.. 2003: Watkinson et al.. 1993). These studies suggest
that these responses to O3 were the result of the rodent hypothermic response, which
serves as a physiological and behavioral defense mechanism to minimize the irritant
effects of O3 inhalation, (Iwasaki et al.. 1998: Arito et al.. 1997). As humans do not
appear to exhibit decreased HR, MAP, and Tco with routine environmental
(Section 6.3.2) or controlled laboratory (Section 6.3.1) exposures to O3, caution must
be used in extrapolating the results of these animal studies to humans.
Other studies have shown that O3 can increase BP in animal models. Rats exposed to
0.6 ppm O3 for 33 days had increased systolic pressure and HR (Revis et al.. 1981).
6-203
-------
Increased BP triggers the release of atrial natriuretic factor (ANF), which has been
found in increased levels in the heart, lungs, and circulation of O3 exposed (0.5 ppm)
rats (Vesely et al., 1994a, b, c). Exposures to high concentrations of O3 (1.0 ppm)
have also been found to lead to heart and lung edema (Friedman et al., 1983), which
could be the result of increased ANF levels. Thus, O3 may increase blood pressure
and HR, leading to increased ANF and tissue edema.
Recent studies report strain differences in HR and FfRV in response to a 2-hour O3
pretreatment followed by exposure to carbon black (CB) in mice (C3H/HeJ [HeJ],
C57BL/6J [B6], and C3H/HeOuJ [OuJ]) (Hamade and Tankerslev. 2009: Hamadeet
al., 2008). These mice strains were chosen from prior studies on lung inflammatory
and hyperpermeability responses to be at increased risk (B6 and OuJ) or resistant
(HeJ) to O3-induced health effects (Kleeberger et al., 2000). HR decreased during O3
pre-exposure for all strains, but recovered during the CB exposure (Hamade et al.,
2008). Percent change in HRV parameters, SDNN (indicating total HRV) and
rMSSD (indicating beat-to-beat HRV), were increased in both C3H mice strains, but
not B6 mice, during O3 pre-exposure and recovered during CB exposure when
compared to the filtered air group. The two C3H strains differ by a mutation in the
Toll-like receptor 4 (TLR4) gene, but these effects did not seem to be related to this
mutation since similar responses were observed. Hamade et al. (2008) speculate that
the B6 and C3H strains differ in mechanisms of HR response after O3 exposure
between withdrawal of sympathetic tone and increase of parasympathetic tone;
however, no direct evidence for this conclusion was reported. The strain differences
observed in HR and HRV suggest that genetic variability affects cardiac responses
after acute air pollutant exposures.
Hamade and Tankersley (2009) continued this investigation of gene-environment
interactions on cardiopulmonary adaptation of O3 and CB induced changes in HR
and HRV using the previously described (Hamade et al., 2008) daily exposure
scheme for 3 consecutive days. By comparing day-1 interim values it is possible to
observe that O3 exposure increased SDNN and rMSSD, but decreased HR in all
strains. Measures of HR and HRV in B6 and HeJ mice recovered to levels consistent
with filtered air treated mice by day 3; however, these responses in OuJ mice
remained suppressed. B6 mice had no change in respiratory rate (RR) after O3
treatment, whereas HeJ mice on days 1 and 2 had increased RR and OuJ mice on
days 2 and 3 exhibited increased RR. VT did not change with treatment among the
strains. Overall, B6 mice were mildly responsive with rapid adaptation, whereas C3
mice were highly responsive with adaptation only in HeJ mice with regards to
changes in cardiac and respiratory responses. HR and HRV parameters were not
equally correlated with VT and RR between the three mice strains, which suggest that
strains vary in the integration of the cardiac and respiratory systems. These complex
interactions could help explain variability in interindividual responses to air
pollution.
Hamade et al. (2010) expanded their investigation to explore the variation of these
strain dependent cardiopulmonary responses with age. As was observed previously,
all experimental mouse strains (B6, HeJ, and OuJ) exhibited decreased HR and
6-204
-------
increased HRV after O3 exposure. Younger O3-exposed mice had a significantly
lower HR compared to older exposed mice, indicating an attenuation of the
bradycardic effect of O3 with age. Younger mice also had a greater increase in
rMSSD in HeJ and OuJ strains and SDNN in HeJ mice. Conversely, B6 mice had a
slightly greater increase in SDNN in aged mice compared to the young mice.
No change was observed in the magnitude of the O3 induced increase of SDNN in
OuJ mice or rMSSD in B6 mice. The B6 and HeJ mice genetically vary in respect to
the nuclear factor erythroid 2-related factor 2 (Nrf-2). The authors propose that the
genetic differences between the mice strains could be altering the formation of ROS,
which tends to increase with age, thus modulating the changes in cardiopulmonary
physiology after O3 exposure.
Strain and age differences in HR and heart function were further investigated in B6
and 129Sl/SvlmJ (129) mice in response to a sequential O3 and filtered air or CB
exposure (Tankersley et al., 2010). Young 129 mice showed a decrease in HR after
O3 or O3 and CB exposure. This bradycardia was not observed in B6 or older
animals in this study, suggesting a possible alteration or adaptation of the autonomic
nervous system activity with age. However, these authors did previously report
bradycardia in similarly aged young B6 mice (Hamade et al., 2010; Hamade and
Tankersley, 2009; Hamade et al., 2008). Ozone exposure in 129 mice also resulted in
an increase in left ventricular chamber dimensions at end diastole (LVEDD) in young
and old mice and a decrease in left ventricular posterior wall thickness at end systole
(PWTES) in older mice. The increase in LVEDD caused a decrease in fractional
shortening, which can be used as a rough indicator of left ventricular function.
Regression analysis revealed a significant interaction between age and strain on HR
and PWTES, which implies that aging affects HR and heart function in response to
O3 differently between mouse strains.
Vascular Disease and Injury
A recent study in young mice (C57B1/6) and rhesus monkeys examined the effects of
short-term O3 exposure (0.5 ppm, 1 or 5 days) on a number of cardiovascular
endpoints (Chuang et al., 2009). Mice exposed to O3 for 5 days had increased HR as
well as mean and diastolic blood pressure. This is in contrast to the bradycardia that
was reported in 18-20 week-old B6 mice treated with O3, as described above
(Hamade and Tankersley, 2009; Hamade et al., 2008). Increased blood pressure
could be explained by the inhibition in endothelial-dependent (acetylcholine)
vasorelaxation from decreased bioavailability of aortic nitric oxide (-NO). Ozone
caused a decrease in aortic NOX (nitrite and nitrate levels) and a decrease in total, but
not phosphorylated, endothelial nitric oxide synthase (eNOS). Ozone also increased
vascular oxidative stress in the form of increased aortic and lung lipid peroxidation
(F2-isoprostane), increased aortic protein nitration (3-nitrotyrosine), decreased aortic
superoxide dismutase (SOD2) protein and activity, and decreased aortic aconitase
activity, indicating specific inactivation by O2~ and ONOO". Mitochondrial DNA
(mtDNA) damage was also used as a measure of oxi dative and nitrative stress in
mice and infant rhesus monkeys exposed to O3. Chuang et al. (2009) observed that
6-205
-------
mtDNA damage accumulated in the lung and aorta of mice after 1 and 5 days of O3
exposure and in the proximal and distal aorta of O3 treated nonhuman primates.
Additionally, genetically hyperlipidemic mice exposed to O3 (0.5 ppm) for 8 weeks
had increased aortic atherosclerotic lesion area (Section 7.3.1), which may be
associated with the short-term exposure changes discussed. Overall, this study
suggests that O3 initiates an oxidative environment by increasing O2~ production,
which leads to mtDNA damage and -NO consumption, known to perturb endothelial
function (Chuang et al.. 2009). Endothelial dysfunction is characteristic of early and
advanced atherosclerosis and coincides with impaired vasodilation and blood
pressure regulation.
Vascular occlusion resulting from atherosclerosis can block blood flow causing
ischemia. The restoration of blood flow in the vessel or reperfusion can cause injury
to the tissue from subsequent inflammation and oxidative damage. Perepu et al.
(2010) observed that O3 exposure (0.8 ppm, 28 or 56 days) enhanced the sensitivity
to myocardial I/R injury in Sprague-Dawley rats while increasing oxidative stress
levels and pro-inflammatory mediators and decreasing production of
anti-inflammatory proteins. Ozone was also found to decrease the left ventricular
developed pressure, rate of change of pressure development, and rate of change of
pressure decay while increasing left ventricular end diastolic pressure in isolated
perfused hearts. In this ex vivo heart model, O3 induced oxidative stress by
decreasing SOD enzyme activity and increasing malondialdehyde levels. Ozone also
elicited a proinflammatory state which was evident by an increase in TNF-a and a
decrease in the anti-inflammatory cytokine IL-10. Perepu et al. (2010) concluded that
O3 exposure may result in a greater I/R injury.
Effects on Cardiovascular-Related Proteins
Increased BP, changes in HRV, and increased atherosclerosis may be related to
increases in the vasoconstrictor peptide, endothelin-1 (amino acids 1-21, ET-1 [1-21]).
Regulation of the pulmonary endothelin system can be affected in rats by inhalation
of PM (0, 5, 50 mg/m3, EHC-93) and O3 (Thomson et al.. 2006: Thomson et al..
2005). Exposure to either O3 (0.8 ppm) or PM increased plasma ET-l[i_2i], ET-3[i_2i],
and the ET-1 precursor peptide, bigET-1. Increases in circulating ET-1 [i_2ij could be
a result of a transient increase in the gene expression of lung preproET-1 and
endothelin converting enzyme-1 (ECE-1) immediately following inhalation of O3 or
PM. These latter gene expression changes (e.g., preproET-1 and ECE-1) were
additive with co-exposure to O3 and PM. Conversely, preproET-3 decreased
immediately after O3 exposure, suggesting the increase in ET-3 [i_2ij was not through
de novo production. A recent study also found increased ET-1 gene expression in the
aorta of O3-exposed rats (Kodavanti et al., 2011). These rats also exhibited an
increase in ETBR after O3 exposure; however, they did not demonstrate increased
biomarkers for vascular inflammation, thrombosis, or oxidation.
Ozone can oxidize protein functional groups and disturb the affected protein. For
example, the soluble plasma protein fibrinogen is oxidized by O3 (0.01-0.03 ppm) in
6-206
-------
vitro, creating fibrinogen and fibrin aggregates, characteristically similar to defective
fibrinogen (Rosenfeld et al., 2009; Rozenfeld et al., 2008). In these studies, oxidized
fibrinogen retained the ability to form fibrin gels that are involved in coagulation,
however the aggregation time increased and the gels were rougher than normal with
thicker fibers. Oxidized fibrinogen also developed the ability to self assemble
creating fibrinogen aggregates that may play a role in thrombosis. Since O3 does not
readily translocate past the ELF and pulmonary epithelium and fibrinogen is
primarily a plasma protein, it is uncertain if O3 would have the opportunity to react
with plasma fibrinogen. However, fibrinogen can be released from the basolateral
face of pulmonary epithelial cells during inflammation, where the deposition of
fibrinogen could lead to lung injury (Lawrence and Simpson-Haidaris. 2004).
Studies on Ozone Reaction Products
Although toxicological studies have demonstrated O3-induced effects on the
cardiovascular system, it remains unclear if the mechanism is through a reflex
response or the result of effects from O3 reaction products (U.S. EPA, 2006b,
1996a). Oxysterols derived from cholesterol ozonation, such as (3-epoxide and
5p,6(3-epoxycholesterol (and its metabolite cholestan-6-oxo-3,5-diol), have been
implicated in inflammation associated with cardiovascular disease (Pulfer et al.,
2005; Pulfer and Murphy, 2004). Two other cholesterol ozonolysis products,
atheronal-A and -B (e.g., cholesterol secoaldehyde), have been found in human
atherosclerotic plaques and shown in vitro to induce foam cell formation and induce
cardiomyocyte apoptosis and necrosis (Sathishkumar et al.. 2005; Wentworth et al..
2003); however, these products have not been found in the lung compartment or
systemically after O3 exposure. The ability to form these cholesterol ozonation
products in the circulation in the absence of O3 exposure complicates their
implication in O3 induced cardiovascular disease.
Although it has been proposed that O3 reaction products released after the interaction
of O3 with ELF constituents (see Section 5.2.3) on O3 interaction with ELF) are
responsible for systemic effects, it is not known whether they gain access to the
vascular space. Alternatively, extrapulmonary release of diffusible mediators, such as
cytokines or endothelins, may initiate or propagate inflammatory responses in the
vascular or systemic compartments (Cole and Freeman, 2009) (Section 5.3.8). Ozone
reacts within the lung to amplify ROS production, induce pulmonary inflammation,
and activate inflammatory cells, resulting in a cascading proinflammatory state and
extrapulmonary release of diffusible mediators that could lead to cardiovascular
injury.
A recent study that examined O3 reaction byproducts has shown that cholesterol
secoaldehyde (e.g., atheronal A) induces apoptosis in vitro in mouse macrophages
(Gao et al.. 2009b) and cardiomyocytes (Sathishkumar et al.. 2009). Additionally,
atheronal-A and -B has been found to induce in vitro macrophage and endothelial
cell proinflammatory events involved in the initiation of atherosclerosis (Takeuchi et
al., 2006). These O3 reaction products when complexed with low density lipoprotein
6-207
-------
upregulate scavenger receptor class A and induce dose-dependent macrophage
chemotaxis. Atheronal-A increases expression of the adhesion molecule, E-selectin,
in endothelial cells, while atheronal-B induces monocyte differentiation. These
events contribute to both monocyte recruitment and foam cell formation in
atherosclerotic vessels. It is unknown whether these O3 reaction products gain access
to the vascular space from the lungs. Alternative explanations include the
extrapulmonary release of diffusible mediators that may initiate or propagate
inflammatory responses in the vascular or systemic compartments.
6-208
-------
Table 6-40
Study3*
Chuang et al. (2009)
Perepuetal. (2010)
Hamade et al.
(2008)
Hamade and
Tankerslev (2009)
Hamade et al.
(2010)
Tankerslev et al.
(2010)
Thomson et al.
(2005)
Thomson et al.
(2006)
Kodavanti et al.
(2011)
Characterization of study details for Section 6.3.3.
Model
Mice; C57BI/6;
M; 6 weeks
Monkey; rhesus
Macaca mulatta',
M; Infant
(180 days old)
Rat; Sprague-
Dawley;
50-75 g
Mice; C57BI/6J,
C3H/HeJ, and
C3H/HeOuJ;
M; 18-20 weeks
Mice; C57BI/6J,
C3H/HeJ, and
C3H/HeOuJ;
M; 18-20 weeks
Mice; C57BI/6J,
C3H/HeJ, and
C3H/HeOuJ; M;
5 or 12 mo old
Mice; C57BI/6J,
12951/SvlmJ;
M/F; 5 or 18 mo old
Rat; Fischer-344;
M; 200-250 g
Rat; Fischer-344;
M; 200-250 g
Rat; Wistar;
M; 10-1 2 weeks
aResults from previous studies are presented
*Studv details for Section 6.3.3.
O3 (ppm)
0.5
0.5
0.8
0.6
(subsequent
CB
exposure,
536 ug/m3)
0.6
(subsequent
CB
exposure,
536 ug/m3)
0.6
(subsequent
CB
exposure,
536 ug/m3)
0.6
(subsequent
CB
exposure,
556 ug/m3)
0.4 or 0.8
0.8
0.5 or 1.0
in Annex Table
Exposure
Duration
1 or 5 days, 8-h/day
5 days, 8-h/day
28 days, 8-h/day
2-h
followed by 3 h of CB
3 days, 2-h/day
followed by 3-h of CB
2-h
followed by 3-h of CB
2-h
followed by 3-h of CB
4-h
4-h
2 days, 5-h/day
Effects
Increased HR and blood pressure.
Initiated an oxidative environment by
increasing vascular O2" production,
which lead to mtDNA damage and -NO
consumption, known to perturb
endothelial function.
Increased aortic mtDNA damage.
Enhanced the sensitivity to myocardial
I/R injury while increasing oxidative
stress and pro-inflammatory mediators
and decreasing production of
anti-inflammatory proteins.
Decreased HR. Strain differences
observed in HRV suggest that genetic
variability affects cardiac responses.
Strains varied in integration of the
cardiac and respiratory systems,
implications in interindividual variability.
B6 mice were mildly responsive with
rapid adaptation, whereas C3 mice were
highly responsive with adaptation only in
HeJ mice with regards to changes in
cardiac and respiratory responses.
Aged mice exhibited attenuated changes
in cardiopulmonary physiology after O3
exposure. Genetic differences between
mice strains could be altering formation
of ROS, which tends to increase with
age, thus modulating O3 induced effects.
Significant interaction between age and
strain on HR and PWTES, which implies
that aging affects the HR and function in
response to O3 differently between
mouse strains.
Activation of the vasoconstricting ET
system. Increased plasma ET-1 through
higher production and slower clearance.
Increased plasma ET-3 not due to de
novo synthesis, unlike ET-1 .
No changes to aortic genes of
thrombosis, inflammation, or proteolysis,
except ET-1 and ETBR (1 .0 ppm).
AX5-14 of the 2006 O3 AQCD and Table 6-23 of the 1996 O3 AQCD.
6-209
-------
Summary of Toxicological Studies
Overall, animal studies suggest that O3 exposure may result in O3 induced
cardiovascular effects. Studies provide evidence for both increased and decreased
HR, however it is uncertain if O3-induced bradycardia would also occur in humans
or if it is due solely to a rodent hypothermic response. Animal studies also provide
evidence for increased HRV, arrhythmias, vascular disease, and injury following
short-term O3 exposure. In addition, a series of studies highlight the role of
gene-environment interactions and age in the induction of effects and attenuation of
responses to O3 exposure.
Biologically plausible mechanisms are present for the cardiovascular effects
observed in animal exposure studies. Further discussion of the modes of action that
may lead to cardiovascular effects can be found in Section 5.3.8. Recent studies
suggest that O3 exposure may disrupt both the NO' and endothelin systems, which
can result in an increase in HR, HRV, and ANF. The observed bradycardia following
O3 exposure may be the result of reflex reactions, including the trigeminocardiac
reflex, evoked following the stimulation of sensory receptors lining the nose and RT.
These mechanisms of parasympathetically-derived cardiac effects are described in
more detail in Section 5.3.2. Additionally, O3 may increase oxidative stress and
vascular inflammation promoting the progression of atherosclerosis and leading to
increased susceptibility to I/R injury. As O3 reacts quickly with the ELF and does not
translocate to the heart and large vessels, studies suggest that the cardiovascular
effects exhibited could be caused by reaction byproducts of O3 exposure. However,
direct evidence of translocation of O3 reaction products to the cardiovascular system
has not been demonstrated in vivo. Alternatively, extrapulmonary release of
diffusible mediators, such as cytokines or endothelins, may initiate or propagate
inflammatory responses in the vascular or systemic compartments leading to the
reported cardiovascular pathologies.
6.3.4 Summary and Causal Determination
In previous O3 reviews (U.S. EPA. 2006b. 1996aX very few studies were described
which examined the effect of short-term O3 exposure on the cardiovascular system.
More recently, the body of scientific evidence available that has examined the effect
of O3 on the cardiovascular system has expanded.
Toxicological studies, although limited in number, provide evidence of O3-induced
cardiovascular effects. These include enhanced I/R injury, disrupted NO-induced
vascular reactivity, decreased cardiac function, increased vascular disease, and
increased HRV following short-term O3 exposure. A number of these effects have
also been observed following long-term O3 exposure (see Section 7.3.1.2). Results of
studies investigating the role of O3 in heart rate regulation are mixed with both
bradycardie and tachycardic responses observed in animal models.
The cardiovascular effects of O3 found in animals may, in part, correspond to
alteration of the autonomic nervous system or to the development and maintenance
6-210
-------
of systemic oxidative stress and a proinflammatory environment that may result from
pulmonary inflammation.
Controlled human exposure studies also suggest cardiovascular effects in response to
short-term O3 exposure and provide some coherence with evidence from animal
toxicology studies. Increases and decreases in high frequency HRV have been
reported following relatively low (120 ppb during rest) and high (300 ppb with
exercise) O3 exposures, respectively. These changes in cardiac function observed in
animal and human studies provide preliminary evidence for O3-induced modulation
of the autonomic nervous system through the activation of neural reflexes in the lung
(see Section 5.3.2). Controlled human exposure studies also support the animal
toxicology studies by demonstrating O3-induced effects on blood biomarkers of
systemic inflammation and oxidative stress as well as changes in biomarkers
suggestive of a prothrombogenic response to O3.
The experimental evidence provides initial biological plausibility for the consistently
positive associations observed in epidemiologic studies of short-term O3 exposure
and cardiovascular mortality. These include studies reviewed in the 2006 O3 AQCD,
recent multicity studies, and the multicontinent APHENA study. The few studies that
examined copollutant confounding found that associations with cardiovascular
mortality remain robust in copollutant models with PM. However, epidemiologic
studies generally do not observe associations between short-term exposure to O3 and
cardiovascular morbidity; studies of cardiovascular-related hospital admissions and
ED visits and other various cardiovascular effects did not find consistent evidence of
a relationship with O3 exposure. The lack of coherence between the results from
studies that examined associations between short-term O3 exposure and
cardiovascular morbidity and cardiovascular mortality complicate the interpretation
of the overall evidence for O3-induced cardiovascular effects.
In conclusion, animal toxicological studies demonstrate O3-induced cardiovascular
effects, and support the strong body of evidence indicating O3-induced
cardiovascular mortality. Animal toxicological and controlled human exposure
studies provide evidence for biologically plausible mechanisms underlying these
O3-induced cardiovascular effects. However, a lack of coherence with epidemiologic
studies of cardiovascular morbidity remains an important uncertainty. Taken
together, the overall body of evidence across disciplines is sufficient to conclude that
there is likely to be a causal relationship between relevant short-term
exposures to O3 and cardiovascular effects.
6.4 Central Nervous System Effects
The 2006 O3 AQCD (U.S. EPA. 2006b) included toxicological evidence indicating
that acute exposures to O3 are associated with alterations in neurotransmitters, motor
activity, short and long term memory, and sleep patterns. Additionally, histological
signs of neurodegeneration have been observed. Reports of headache, dizziness, and
irritation of the nose with O3 exposure are common complaints in humans, and some
6-211
-------
behavioral changes in animals may be related to these symptoms rather than
indicative of neurotoxicity. Peterson and Andrews (1963) and Tepper et al. (1983)
showed that mice would alter their behavior to avoid O3 exposure. Murphy et al.
(1964) and Tepper et al. (1982) showed that running-wheel behavior was suppressed,
and Tepper et al. (1985) subsequently demonstrated the effects of a 6-hour exposure
to O3 on the suppression of running-wheel behavior in rats and mice, with the lowest
effective concentration being about 0.12 ppm O3 in the rat and about 0.2 ppm in the
mouse. The suppression of active behavior by 6 hours of exposure to 0.12 ppm O3
has recently been confirmed by Martrette et al. (2011) in juvenile female rats, and the
suppression of three different active behavior parameters was found to become more
pronounced after 15 days of exposure. A table of studies examining the effects of O3
on behavior can be found on p 6-128 of the 1996 O3 AQCD. Generally speaking,
transient changes in behavior in rodent models appear to be dependent on a complex
interaction of factors such as (1) the type of behavior being measured, with some
behaviors increased and others suppressed; (2) the factors motivating that behavior
(differences in reinforcement); and (3) the sensitivity of the particular behavior
(e.g., active behaviors are more affected than more sedentary behaviors). Many
behavioral changes are likely to result from avoidance of irritation, but more recent
studies indicate that O3 also directly affects the CNS.
Research in the area of O3-induced neurotoxicity has notably increased over the past
few years, with the majority of the evidence coming from toxicological studies that
examined the association between O3 exposure, neuropathology, and
neurobehavioral effects. As discussed below, these studies demonstrated that
exposure to O3 can produce a variety of CNS effects including behavioral deficits,
morphological changes, and oxidative stress in the brains of rodents.In these rodent
studies, interestingly, CNS effects were reported at O3 concentrations that were
generally lower than those concentrations commonly observed to produce pulmonary
or cardiac effects in rats. A recent epidemiologic study provides new evidence, which
is coherent with the toxicological evidence indicating that ambient O3 exposure may
result in cognitive function decrements. This study is discussed in detail in Chapter 7
(Section 7.5.1) because it focuses on long-term exposures to O3.
A number of new studies demonstrate various perturbations in neurologic function or
histology, including changes similar to those observed with Parkinson's and
Alzheimer's disease pathologies occurring in similar regions of the brain
(Table 6-41). Many of these include exposure durations ranging from short-term to
long-term, and as such are discussed here and in Chapter 7 with emphasis on the
effects resulting from exposure durations relevant to the respective chapter. Several
studies assess short- and long-term memory acquisition via passive avoidance
behavioral testing and find decrements in test performance after O3 exposure,
consistent with the aforementioned observation made in humans by Chen and
Schwartz (2009). Impairment of long-term memory has been previously described in
rats exposed to 0.2 ppm O3 for 4 hours (Rivas-Arancibia et al., 1998) and in other
studies of 4-hour exposures at concentrations of 0.7 to 1 ppm (Dorado-Martinez et
al.,2001; Rivas-Arancibia et al., 2000; Avila-Costa et al., 1999). More recently,
6-212
-------
statistically significant decreases in both short and long-term memory were observed
in rats after 15 days of exposure to 0.25 ppm O3 (Rivas-Arancibia et al., 2010).
The central nervous system is very sensitive to oxidative stress, due in part to its high
content of polyunsaturated fatty acids, high rate of oxygen consumption, and low
antioxidant enzyme capacity. Oxidative stress has been identified as one of the
pathophysiological mechanisms underlying neurodegenerative disorders such as
Parkinson's and Alzheimer's disease, among others (Simonian and Covle, 1996). It is
also believed to play a role in altering hippocampal function, which causes cognitive
deficits with aging (Vanguilder and Freeman, 2011). A particularly common finding
in studies of O3-exposed rats is lipid peroxidation in the brain, especially in the
hippocampus, which is important for higher cognitive function including contextual
memory acquisition. Performance in passive avoidance learning tests is impaired
when the hippocampus is injured, and the observed behavioral effects are well
correlated with histological and biochemical changes in the hippocampus, including
reduction in spine density in the pyramidal neurons (Avila-Costa et al., 1999),
lipoperoxidation (Rivas-Arancibia et al., 2010; Dorado-Martinez et al., 2001),
progressive neurodegeneration, and activated and phagocytic microglia (Rivas-
Arancibia et al., 2010). The hippocampus is also one of the main regions affected by
age-related neurodegenerative diseases, including Alzheimer's disease, and it may be
more sensitive to oxidative damage in aged rats. In a study of young (47 days) and
aged (900 days) rats exposed to 1 ppm O3 for 4 hours, O3-induced lipid peroxidation
occurred to a greater extent in the striatum of young rats, whereas it was highest in
the hippocampus in aged rats (Rivas-Arancibia et al., 2000). Martinez-Canabal and
Angora-Perez (2008) showed exposure of rats to 0.25 ppm, 4h/day, for 7, 15, or
30 days increased lipoperoxides in the hippocampus. This effect was observed at day
7 and continued to increase with time, indicating cumulative oxidative damage.
O3-induced changes in lipid peroxidation, neuronal death, and COX-2 positive cells
in the hippocampus could be significantly inhibited by daily treatment with growth
hormone (GH), which declines with age in most species. The protective effect of GH
on -induced oxidative stress was greatest at 15 days of exposure and was non-
significant at day 30. Consistent with these findings, lipid peroxidation in the
hippocampus of rats was observed to increase significantly after a 30-day exposure to
0.25 ppm, but not after a single 4-hour exposure to the same concentration (Mokoena
et al.. 2010). However, 4 hours of exposure was sufficient to cause significant
increases in lipid peroxidation when the concentration was increased to 0.7 ppm, and
another study observed lipid peroxidation after a 4-hour exposure to 0.4 ppm
(Dorado-Martinez et al.. 2001).
Other commonly affected areas of the brain include the striatum, substantia nigra,
cerebellum, olfactory bulb, and frontal/prefrontal cortex. The striatum and substantia
nigra are particularly sensitive to oxidative stress because the metabolism of
dopamine, central to their function, is an oxidative process perturbed by redox
imbalance. Oxidative stress has been implicated in the premature death of substantia
nigra dopamine neurons in Parkinson's disease. Angoa-Perez et al. (2006) have
shown progressive lipoperoxidation in the substantia nigra and a decrease in nigral
dopamine neurons in ovariectomized female rats exposed to 0.25 ppm O3, 4h/day,
6-213
-------
for 7, 15, or 30 days. Estradiol, an antioxidant, attenuated O3-induced oxidative
stress and nigral neuronal death, and the authors note that in humans, estrogen
therapy can ameliorate symptoms of Parkinson's disease, which is more prevalent in
men. Progressive oxidative stress has also been observed in the striatum and
substantia nigra of rats after 15 and 30 days of exposure to 0.25 ppm O3 for
4 hours/day, along with a loss of dopaminergic neurons from the substantia nigra
(Perevra-Mufioz et al.. 2006). Decreases in motor activity were also observed at 15
and 30 days of exposure, consistent with other reports flVIartrette et al.. 2011:
Dorado-Martinez et al.. 2001). Using a similar O3 exposure protocol, Santiago-Lopez
et al. (2010) also observed a progressive loss of dopaminergic neurons within the
substantia nigra, accompanied by alterations in the morphology of remaining cells
and an increase in p53 levels and nuclear translocation.
The olfactory bulb also undergoes oxidative damage in O3 exposed animals, in some
cases altering olfactory-dependent behavior. Lipid peroxidation was observed in the
olfactory bulbs of ovariectomized female rats exposed to 0.25 ppm O3 (4 hours/day)
for 30 or 60 days (Guevara-Guzman et al., 2009). Ozone also induced decrements in
a selective olfactory recognition memory test, and the authors note that early deficits
in odor perception and memory are components of human neurodegenerative
diseases. The decrements in olfactory memory were not due to damaged olfactory
perception based on other tests. However, deficits in olfactory perception emerged
with longer exposures (discussed in Chapter 7). As with the study by Angoa-Perez et
al. (2006) described above, a protective effect for estradiol was demonstrated for
both lipid peroxidation and olfactory memory defects. The role of oxidative stress in
memory deficits and associated morphological changes has also been demonstrated
via attenuation by other antioxidants as well, such as a-tocopherol (Guerrero et al.,
1999) and taurine (Rivas-Arancibia et al.. 2000).It is unclear how persistent these
effects might be. One study of acute exposure, using 1 ppm O3 for 4 hours, observed
morphological changes in the olfactory bulb of rats at 2 hours, and 1 and 10 days, but
not 15 days, after exposure (Colin-Barenque et al.. 2005).
Other acute studies also report changes in the CNS. Lipid peroxidation was observed
in multiple regions of the brain after a 1- to 9-hour exposure to 1 ppm O3 (Escalante-
Membrillo et al., 2005). Ozone has also been shown to alter gene expression of
endothelin-1 (pituitary) and inducible nitric oxide synthase (cerebral hemisphere)
after a single 4-hour exposure to 0.8 ppm O3, indicating potential cerebrovascular
effects. This concentration-dependent effect was not observed at 0.4 ppm O3
(Thomson et al., 2007). Vascular endothelial growth factor was upregulated in
astroglial cells in the central respiratory areas of the brain of rats exposed to 0.5 ppm
O3 for 3 hours (Araneda et al., 2008). The persistence of CNS changes after a single
exposure was also examined and the increase in vascular endothelial growth factor
was present after a short (3 hours) recovery period. Thus, there is evidence that
O3-induced CNS effects are both concentration- and time-dependent.
Because O3 can produce a disruption of the sleep-wake cycle (U.S. EPA. 2006b).
Alfaro-Rodriguez and Gonzalez-Pifia (2005) examined whether acetylcholine in a
region of the brain involved in sleep regulation was altered by O3. After a 24-hour
6-214
-------
exposure to 0.5 ppm O3, the acetylcholine concentration in the medial preoptic area
was decreased by 58% and strongly correlated with a disruption in paradoxical sleep.
Such behavioral-biochemical effects of O3 are confirmed by a number of studies
which have demonstrated morphological and biochemical changes in rats.
CNS effects have also been demonstrated in newborn and adult rats whose only
exposure to O3 occurred in utero. Several neurotransmitters were assessed in male
offspring of dams exposed to 1 ppm O3 during the entire pregnancy (Gonzalez-Pina
et al., 2008). The data showed that catecholamine neurotransmitters were affected to
a greater degree than indole-amine neurotransmitters in the cerebellum. CNS
changes, including behavioral, cellular, and biochemical effects, have also been
observed after in utero exposure to 0.5 ppm O3 for 12 hours/day from
gestational days 5-20 (Boussouar et al., 2009). Tyrosine hydroxylase labeling in the
nucleus tractus solatarius was increased after in utero exposure to O3 whereas Fos
protein labeling did not change. When these offspring were challenged by
immobilization stress, neuroplasticity pathways, which were activated in air-exposed
offspring, were inhibited in O3-exposed offspring. Although an O3 exposure C-R
was not studied in these two in utero studies, it has been examined in one study.
Santucci et al. (2006) investigated behavioral effects and gene expression after in
utero exposure of mice to as little as 0.3 ppm O3. Increased defensive/submissive
behavior and reduced social investigation were observed in both the 0.3 and 0.6 ppm
O3 groups. Changes in gene expression of brain-derived neurotrophic factor (BDNF,
increased in striatum) and nerve growth factor (NGF, decreased in hippocampus)
accompanied these behavioral changes. Thus, these three studies demonstrate that
CNS effects can occur as a result of in utero exposure to O3, and although the mode
of action of these effects is not known, it has been suggested that circulating lipid
peroxidation products may play a role (Boussouar et al.. 2009). Importantly, these
CNS effects occurred in rodent models after in utero only exposure to relevant
concentrations of O3.
6-215
-------
Table 6-41 Central nervous system and behavioral effects of short-term
exposure in rats.
Study
Martrette et al. (2011)
Angoa-Perez et al. (2006)
Guevara-Guzman et al.
(2009)
Martinez-Canabal and
Angora-Perez (2008)
Perevra-Munoz et al. (2006)
Rivas-Arancibia et al. (2010)
Santiago-Lopez et al. (2010)
Thomson et al. (2007)
Alfaro-Rodriguez and
Gonzalez-Pina (2005)
Araneda et al. (2008)
Model
Rat; Wistar;
F; Weight: 152g;
7 weeks old
Rat; Wistar;
F; Weight: 300g;
ovariectomized
Rat; Wistar;
F; 264g;
ovariectomized
Rat; Wistar;
M; Weight: 300g
Rat; Wistar;
M; 250-300g
Rat; Wistar;
M; 250-300g
Rat; Wistar;
M; 250-300g
Rat; Fischer-344;
M; 200-250g
Rat; Wistar;
M; 292g
Rats; Sprague-
Dawley; M; 280-
320g
03
(ppm)
0.12
0.25
0.25
0.25
0.25
0.25
0.25
0.4; 0.8
0.5
0.5
Exposure
Duration
1-15 days,
6 h/day
7 to 60 days,
4-h/day,
5 days/week
30 and 60 days,
4h/day
7 to 30 days,
4-h/day
15 and 30 days,
4-h/day
1 5 to 90 days,
4-h/ day
15, 30, and 60
days, 4-h/day
4-h;
assays at 0 and
24 h postexposure
24-h
3-h
(measurements
taken at 0 h and 3
h after exposure)
Effects
Significant decrease in rearing, locomotor
activity, and jumping activity at day 1 , with a
further decrease in these activities by day
15.
Progressive lipid peroxidation and loss of
tyrosine hydrolase-immunopositive neurons
in the substantia nigra starting at 7 days.
Estradiol treatment protected against lipid
peroxidation and decreases in estrogen
receptors and dopamine p-hydroxylase in
olfactory bulbs along with deficits in olfactory
recognition memory.
Growth hormone inhibited O3-induced
increases in lipoperoxidation and COX-2
positive cells in the hippocampus.
Decreased motor activity, increased lipid
peroxidation, altered morphology, and loss
of dopamine neurons in substantia nigra and
striatum, increased expression of DARPP-
32, iNOS, and SOD.
Ozone produced significant increases in lipid
peroxidation in the hippocampus, and
altered the number of p53 positive
immunoreactive cells, activated and
phagocytic microglia cells, GFAP
immunoreactive cells, and doublecortine
cells, and short- and long-term memory-
retention latency.
Progressive loss of dopamine reactivity in
the substantia nigra, along with
morphological changes. Increased p53
levels and nuclear translocation.
At 0.8 ppm, O3 produced rapid perturbations
in the ET-NO pathway gene expression in
the brain. Ozone induced a small but
significant time- and concentration-
dependent increase in prepro-endothelin-1
mRNA levels in the cerebral hemisphere and
pituitary, whereas TNFa and iNOS mRNA
levels were decreased at 0 h and
unchanged or increased, respectively, at 24
h.
During the light phase, O3 caused a
significant decrease in paradoxical sleep
accompanied by a significant decrease in
Ach levels in the hypothalamic medial
preoptic area. The same effects occurred
during the dark phase exposure to O3 in
addition to a significant increase in slow-
wave sleep and decrease in wakefulness.
Ozone upregulated VEGF in astroglial cells
located in the respiratory center of the brain.
VEGF co-located with IL-6 and TNF in cells
near blood vessel walls, and blood vessel
area was markedly increased.
6-216
-------
Exposure
Study
Boussouar et al. (2009)
Soulage et al. (2004)
Calderon Guzman et al.
(2006): Calderon Guzman et
al. (2005)
Colin-Barengue et al. (2005)
Escalante-Membrillo et al.
(2005)
Gonzalez-Pina et al. (2008)
Model
Rat; Sprague-
Dawley; M; adult
offspring of
prenatally exposed
dams; 403-41 4g
Rat; Sprague-
Dawley; M; Approx.
7 weeks old
Rat; Wistar;
M;21 days old;
well-nourished and
malnourished
groups
Rats; Wistar; M;
250-300g
Rats; Wistar;
M; 280-320g
Rat; Wistar;
M;
(ppm) Duration
0.5 From embryonic
day E5 to E20 for
1 -h/day;
immobilization
stress
0.7 5-h
0.75 15 successive
days for 4-h/day
1 .0 4-h; assays
at2-h, 24-h, 10
days, and 15 days
after exposure
1.0 1 -, 3-, 6-, or 9-h
1 12-h/day,
21 days of
gestation; assays
at 0, 5, & 10 days
postnatal
Effects
Prenatal O3 exposure had a long term
impact on the nucleus tractus solitarius of
adult rats, as revealed during immobilization
stress.
Ozone produced differential effects on
peripheral and central components of the
sympatho-adrenal system. While
catecholamine biosynthesis was increased
in portions of the brain, the catecholamine
turnover rate was significantly increased in
the heart and cerebral cortex and inhibited in
the lung and striatum.
A significant decrease in body weight was
observed in both well nourished (WN) and
malnourished (MN) rats after O3 exposure.
Localized ATPase, TEARS, and GSH levels
changed in response to O3 in certain brain
areas and the O3-induced changes were
dependent on nutritional condition.
A significant loss of dendritic spines in
granule cells of the olfactory bulb occurred
at 2 hrs to 10 days after exposure.
Cytological and ultrastructural changes
returned toward normal morphology by 15
days.
Significant increases in TEARS occurred in
hypothalamus, cortex, striatum, midbrain,
thalamus, and pons. Partial but significant
recovery was observed by 3 h after the 9 h
exposure.
Prenatal O3 exposure produced significant
decreases in cerebellar monoamine but not
indolamine content at 0 and 5 days after
birth with a partial recovery by 10 days. 5-
hydroxy-indole-acetic acid levels were
significantly increased at 10 days.
6.4.1 Neuroendocrine Effects
According to the 2006 O3 AQCD, early studies suggested an interaction of O3 with
the pituitary-thyroid-adrenal axis, because thyroidectomy, hypophysectomy, and
adrenalectomy protected against the lethal effects of O3. Concentrations of
0.7-1.0 ppm O3 for a 1-day exposure in male rats caused changes in the parathyroid,
thymic atrophy, decreased serum levels of thyroid hormones and protein binding, and
increased prolactin. Increased toxicity to O3 was reported in hyperthyroid rats and T3
supplementation was shown to increase metabolic rate and pulmonary injury in the
lungs of O3-treated animals. The mechanisms by which O3 affects neuroendocrine
function are not well understood, but previous work suggests that high ambient levels
of O3 can produce marked neural disturbances in structures involved in the
integration of chemosensory inputs, arousal, and motor control, effects that may be
6-217
-------
responsible for some of the behavioral effects seen with O3 exposure. A more recent
study exposing immature female rats to 0.12 ppm O3 demonstrated significantly
increased serum levels of the thyroid hormone free T3 after 15 days of exposure,
whereas free T4 was unchanged (Martrette et al.. 2011). These results are in contrast
to those previously presented whereby 1 ppm O3 for 1 day significantly decreased T3
and T4 (demons and Garcia. 1980). although comparisons are made difficult by
highly disparate exposure regimens along with sex differences. Martrette et al.
(2011) also demonstrated significantly increased corticosterone levels after 15 days
of exposure, suggesting a stress related response.
6.4.2 Summary and Causal Determination
In rodents, O3 exposure has been shown to cause physicochemical changes in the
brain indicative of oxidative stress and inflammation. Newer toxicological studies
add to earlier evidence that acute exposures to O3 can produce a range of effects on
the central nervous system and behavior. Previously observed effects, including
neurodegeneration, alterations in neurotransmitters, short and long term memory, and
sleep patterns, have been further supported by recent studies. In instances where
pathology and behavior are both examined, animals exhibit decrements in behaviors
tied to the brain regions or chemicals found to be affected or damaged. For example,
damage in the hippocampus, which is important for memory acquisition, was
correlated with impaired performance in tests designed to assess memory. Thus the
brain is functionally affected by O3 exposure. The single epidemiologic study
conducted showed an association between O3 exposure and memory deficits in
humans as well, albeit on a long-term exposure basis. Notably, exposure to O3 levels
as low as 0.25 ppm for 7 days has resulted in progressive neurodegeneration and
deficits in both short and long-term memory in rodents. Examination of changes in
the brain at lower exposure concentrations or at 0.25 ppm for shorter durations has
not been reported, but 0.12 ppm O3 has been shown to alter behavior. It is possible
that some behavioral changes may reflect avoidance of irritation as opposed to
functional changes in brain morphology or chemistry, but in many cases functional
changes are related to oxidative stress and damage. In some instances, changes were
dependent on the nutritional status of the rats (high versus low protein diet). For
example, O3 produced an increase in glutathione in the brains of rats fed the high
protein diet but decreases in glutathione in rats fed low protein chow (Calderon
Guzman et al.. 2006). The hippocampus, one of the main regions affected by age-
related neurodegenerative diseases, appears to be more sensitive to oxidative damage
in aged rats (Rivas-Arancibia et al.. 2000). and growth hormone, which declines with
age in most species, may be protective (Martinez-Canabal and Angora-Perez. 2008).
Developing animals may also be sensitive, as changes in the CNS, including
biochemical, cellular, and behavioral effects, have been observed in juvenile and
adult animals whose sole exposure occurred in utero, at levels as a low as 0.3 ppm.
A number of studies demonstrate O3-induced changes that are also observed in
human neurodegenerative disorders such as Alzheimer's and Parkinson's disease,
including signs of oxidative stress, loss of neurons/neuronal death, reductions in
6-218
-------
dopamine levels, increased COX-2 expression, and increases in activated microglia
in important regions of the brain (hippocampus, substantia nigra).
Thus, evidence for neurological effects from epidemiologic and controlled human
exposure studies is lacking. However, the toxicological evidence for the impact of O3
on the brain and behavior is strong, and suggestive of a causal relationship
between O3 exposure and effects on the central nervous system.
6.5 Effects on Other Organ Systems
6.5.1 Effects on the Liver and Xenobiotic Metabolism
Early investigations of the effects of O3 on the liver centered on xenobiotic
metabolism, and the prolongation of drug-induced sleeping time, which was
observed at 0.1 ppm O3 (Graham et al.. 1981). In some species, only adults and
especially females were affected. In rats, high (1.0-2.0 ppm for 3 hours) acute O3
exposures caused increased production of NO by hepatocytes and enhanced protein
synthesis (Laskin et al.. 1996: Laskin et al.. 1994). Except for the earlier work on
xenobiotic metabolism, the responses occurred only after very high acute O3
exposures. One study, conducted at 1 ppm O3 exposure, has been identified (Last et
al.. 2005) in which alterations in gene expression underlying O3-induced cachexia
and downregulation of xenobiotic metabolism were examined. A number of the
downregulated genes are known to be interferon (IFN) dependent, suggesting a role
for circulating IFN. A more recent study by Aibo et al. (2010) demonstrates
exacerbation of acetaminophen-induced liver injury in mice after a single 6-hour
exposure to 0.25 or 0.5 ppm O3. Data indicate that O3 may worsen drug-induced
liver injury by inhibiting hepatic repair. The O3-associated effects shown in the liver
are thought to be mediated by inflammatory cytokines or other cytotoxic mediators
released by activated macrophages or other cells in the lungs (Laskin and Laskin.
2001: Laskin et al.. 1998: Vincent et al.. 1996a). Recently, increased peroxidated
lipids were detected in the plasma of O3 exposed animals (Santiago-Lopez et al.,
2010).
In summary, mediators generated by O3 exposure may cause effects on the liver in
laboratory rodents. Ozone exposures as low as 0.1 ppm have been shown to affect
drug-induced sleeping time, and exposure to 0.25 ppm can exacerbate liver injury
induced by a common analgesic. However, very few studies at relevant
concentrations have been conducted, and no data from controlled human exposure or
epidemiologic studies are currently available. Therefore the collective evidence is
inadequate to determine if a causal relationship exists between short-term O3
exposure and effects on the liver and metabolism.
6-219
-------
6.5.2 Effects on Cutaneous and Ocular Tissues
In addition to the lungs, the skin is highly exposed to O3 and contains O3 reactive
targets (polyunsaturated fatty acids) that can produce lipid peroxides. The 2006 O3
AQCD (U.S. EPA, 2006b) reported that although there is evidence of oxidative stress
at near ambient O3 concentrations, skin and eyes are only affected at high
concentrations (greater than 1-5 ppm). Ozone exposure (0.8 ppm for 7 days) induces
oxi dative stress in the skin of hairless mice, along with proinflammatory cytokines
(Valacchi et al, 2009). A recent study demonstrated that 0.25 ppm O3 differentially
alters expression of metalloproteinases in the skin of young and aged mice,
indicating that age may potentially increase risk of oxidative stress (Fortino et al.,
2007). In young mice, healing of skin wounds is not significantly affected by O3
exposure (Lim et al., 2006). However, exposure to 0.5 ppm O3 for 6 hours/day
significantly delays wound closure in aged mice. As with effects on the liver
described above, the effects of O3 on the skin and eyes have not been widely studied,
and information from controlled human exposure or epidemiologic studies is not
currently available. Therefore the collective evidence is inadequate to determine
if a causal relationship exists between short-term O3 exposure and effects on
cutaneous and ocular tissues.
6.6 Mortality
6.6.1 Summary of Findings from 2006 O3 AQCD
The 2006 O3 AQCD (U.S. EPA. 2006b) reviewed a large number of time-series
studies consisting of single- and multicity studies, and meta-analyses. In the large
U.S. multicity studies that examined all-year data, summary effect estimates
corresponding to single-day lags ranged from a 0.5-1% increase in all-cause
(nonaccidental) mortality per a standardized unit increase in O3 of 20 ppb for
24-h avg, 30 ppb for 8-h max, and 40 ppb for 1-h max as discussed in Section 2.5.
The association between short-term O3 exposure and mortality was substantiated by
a collection of meta-analyses and international multicity studies. The studies
evaluated found some evidence for heterogeneity in O3 mortality risk estimates
across cities and studies. Studies that conducted seasonal analyses, although more
limited in number, reported larger O3 mortality risk estimates during the warm or
summer season. Overall, the 2006 O3 AQCD identified robust associations between
various measures of daily ambient O3 concentrations and all-cause mortality, with
additional evidence for associations with cardiovascular mortality, which could not
be readily explained by confounding due to time, weather, or copollutants. However,
it was noted that multiple uncertainties remain regarding the O3-mortality
relationship including: the extent of residual confounding by copollutants; factors
that modify the O3-mortality association; the appropriate lag structure for identifying
O3-mortality effects (e.g., single-day lags versus distributed lag model); the shape of
6-220
-------
the O3-mortality C-R function and whether a threshold exists; and the identification
of populations at-risk to O3-related health effects. Collectively, the 2006 O3 AQCD
concluded that "the overall body of evidence is highly suggestive that O3 directly or
indirectly contributes to non-accidental and cardiopulmonary-related mortality."
6.6.2 Associations of Mortality and Short-Term Os Exposure
Recent studies that examined the association between short-term O3 exposure and
mortality further confirmed the associations reported in the 2006 O3 AQCD. New
multicontinent and multicity studies reported consistent positive associations
between short-term O3 exposure and all-cause mortality in all-year analyses, with
additional evidence for larger mortality risk estimates during the warm or summer
months (Figure 6-27 [and Table 6-421). These associations were reported across a
range of ambient O3 concentrations that were in some cases quite low (Table 6-43).
Study
Grypariset al. (2004)
Bell etal. (2007)
Schwartz (2005)
Bell and Dominici (2008)
Bell etal. (2004)a
Levy etal. (2005 )a
Katsouyanni etal. (2009)
Bell etal. (2005)a
ltd etal. (2005)a
Wonget al. (2010)
Katsouyanni etal. (2009)
Cakmaketal. (2011)
Katsouyanni etal. (2009)
Katsouyanni etal. (2009)b
Samoli etal. (2009)
Bell etal. (2004)a
Schwartz (2005)
Zanobetti and Schwartz (2008)
Zanobetti and Schwartz (2008)
Franklin and Schwartz (2008)
Grypariset al. (2004)
Medina-Ramon and Schwartz (2008)
Katsouyanni etal. (2009)
Bell etal. (2005)a
Katsouyanni etal. (2009
Katsouyanni etal. (2009 b
Levy etal. (2005 )a
ltd etal. (2005)a
Katsouyanni etal. (2009)
Stafdggia etal. (2010)
Location
APHEA2 (23 cities)
98 U.S. communities
14 U.S. cities
98 U.S. communjtjes
95 U.S. communities
U.S. and Non-U.S.
APHENA-Europe
U.S. and Non-llS.
U.S. and Non-U.S.
.. ..
PAP A 14 cities)
APHENA-U.S.
7 Chilean cities
APHENA-Canada
APHENA-Canada
21 European cities
95 U.S. communities
14 U.S. cities
48 U.S. cities
48 U.S. cities
18 U.S. communities
APHEA2 (21 cities)
48 U.S. cities
APHENA-Europe
U.S. and Non-llS.
APHENA-Canada
APHENA-Canada
U.S. and Non-U.S.
U.S. and Non-U.S.
APHENA-U.S.
10 Italian cities
Lag
o-i
0-1
0
0-6
0-6
DL(0-2)
0-1
DL
0-2
0-1
0-6
0
0
0-3
0
0-1
0-2
DL(0-2)
DL(0-2)
DL 0-2
DLjO-2)
DL(0-5)
All-Year
Summer
1357
% Increase
11
Note: Effect estimates are for a 40 ppb increase in 1-h max, 30 ppb increase in 8-h max, and 20 ppb increase in 24-h avg O3
concentrations. An "a" represent multicity studies and meta-analyses from the 2006 O3 AQCD. Bell et al. (2005). Ito et al. (2005).
and Lew et al. (2005) used a range of lag days in the meta-analysis: Lag 0,1,2, or average 0-1 or 1 -2; single-day lags from 0 to
3; and lag 0 and 1-2; respectively. A "b" represents risk estimates from APHENA-Canada standardized to an approximate IQR of
5.1 ppb fora 1-h max increase in O3 concentrations (see explanation in Section 6.2.7.2).
Figure 6-27 Summary of mortality risk estimates for short-term O3 exposure
and all-cause (nonaccidental) mortality from all-year and summer
season analyses.
6-221
-------
Table 6-42 Corresponding effect estimates for Figure 6-27.
Study*
Location
Lag
Avg Time
% Increase (95% Cl)
All-year
Grypariset al. (2004)
Bell et al. (2007)
Schwartz (2005a)
Bell and Dominici (2008)
Bell et al. (2004)a
Lew et al. (2005)a
Katsouvanni et al. (2009)
Bell et al. (2005)a
Ito et al. (2005)a
Wongetal. (2010)
Katsouvanni et al. (2009)
Cakmaketal. (2011)
Katsouvanni et al. (2009)
Katsouvanni etal. (2009)"
APHEA2 (23 cities)
98 U.S. communities
14 U.S. cities
98 U.S. communities
95 U.S. communities
U.S. and Non-U.S.
APHENA-europe
U.S. and Non-U.S.
U.S. and Non-U.S.
PAPA (4 cities)
APHENA-U.S.
7 Chilean cities
APHENA-Canada
APHENA-Canada
0-1
0-1
0
0-6
0-6
DL(0-2)
—
—
0-1
DL(0-2)
DL(0-6)
DL(0-2)
DL(0-2)
1-h max
24-h avg
1-h max
24-h avg
24-h avg
24-h avg
1-h max
24-h avg
24-h avg
8-h avg
1-h max
8-h max
1-h max
1-h max
0.24 (-0.86, 1 .98)
0.64 (0.34, 0.92)
0.76(0.13, 1.40)
1.04(0.56, 1.55)
1 .04 (0.54, 1 .55)
1.64(1.25,2.03)
1.66(0.47,2.94)
1.75(1.10,2.37)
2.20 (0.80, 3.60)
2.26(1.36,3.16)
3.02(1.10,4.89)
3.35(1.07,5.75)
5.87(1.82,9.81)
0.73 (0.23, 1 .20)
Summer
Samoli et al. (2009)
Bell et al. (2004)a
Schwartz (2005a)
Zanobetti and Schwartz (2008a)
Zanobetti and Schwartz (2008b)
Franklin and Schwartz (2008)
Grypariset al. (2004)
Medina-Ramon and Schwartz (2008)
Katsouvanni etal. (2009)
Bell et al. (2005)a
Katsouvanni et al. (2009)
Katsouvanni etal. (2009)
Lew et al. (2005)a
Ito et al. (2005)a
Katsouvanni etal. (2009)
Stafoggia etal. (2010)
'Studies included from Figure 6-27.
aMulticity studies and meta-analyses from the
21 European cities
95 U.S. communities
14 U.S. cities
48 U.S. cities
48 U.S. cities
18 U.S. communities
APHEA2 (21 cities)
48 U.S. cities
APHENA-europe
U.S. and Non-U.S.
APHENA-Canada
APHENA-Canada
U.S. and Non-U.S.
U.S. and Non-U.S.
APHENA-U.S.
10 Italian cities
2006 03AQCD. Bell etal.
0-1
0-6
0
0
0-3
0
0-1
0-2
DL(0-2)
DL(0-2)
DL(0-2)
DL(0-2)
DL(0-5)
(2005)a, Ito
8-h max
24-h avg
1-h max
8-h max
8-h max
24-h avg
8-h max
8-h max
1-h max
24-h avg
1-h max
1-h max
24-h avg
24-h avg
1-h max
8-h max
et al. (2005)a, and
0.66 (0.24, 1 .05)
0.78 (0.26, 1 .30)
1.00(0.30, 1.80)
1.51 (1.14, 1.87)
1.60(0.84,2.33)
1.79(0.90,2.68)
1.80(0.99,3.06)
1.96(1.14,2.82)
2.38(0.87,3.91)
3.02(1.45,4.63)
3.34(1.26,5.38)
0.42(0.16,0.67)
3.38 (2.27, 4.42)
3.50(2.10,4.90)
3.83(1.90,5.79)
9.15(5.41,13.0)
Lew et al. (2005)a used a
range of lag days in the meta-analysis: Lag 0, 1, 2, or average 0-1 or 1 -2; Single-day lags from 0-3; and Lag 0 and 1 -2;
respectively.
bRisk estimates from APHENA-Canada standardized to an approximate IQR of 5.1 ppb for a 1-h max increase in O3 concentrations
(see explanation in Section 6.2.7.2).
6-222
-------
Table 6-43
Study
Gryparis et al.
(2004)"
Schwartz
(2005a)b
Bell et al.
(2004)
Bell et al.
(2007)
Bell and
Dominici
(2008)
Franklin and
Schwartz
(2008)
Katsouvanni et
al. (2009)"'e
Medina-Ramon
and Schwartz
(2008)"
Samoli et al.
(2009)"
Stafoggia et al.
(2010)
Cakmaket al.
(2011)
Wong et al.
(2010)
Zanobetti and
Schwartz
(2008b)
Range of mean and upper percentile O3 concentrations in previous
and recent multicity studies.
Location
23 European
cities
(APHEA2)
14 U.S. cities
95 U.S.
communities
(NMMAPS)
98 U.S.
communities
(NMMAPS)
98 U.S.
communities
(NMMAPS)
18 U.S.
communities
NMMAPS
12 Canadian
cities
(APHEA2)
48 U.S. cities
21 European
cities
(APHEA2)
10 Italian cities
7 Chilean
cities
PAPA (4 cities)
48 U.S. cities
Years
1990-1997
1986-1993
1987-2000
1987-2000
1987-2000
(All year and
May-Sept)
2000-2005
(May-Sept)
1987-1996
(Canada and
U.S.) varied by
city for Europe
1989-2000
(May-Sept)
1990-1997
(June-Aug)
2001 -2005
(Apr-Sept)
1997-2007
1999-2003
(Bangkok)
1996-2002
(Hong Kong)
2001 -2004
(Shanghai)
2001 -2004
(Wuhan)
1989-2000
(June-Aug)
Averaging
Time
1-h max
8-h max
1-h max
24-h avg
24-h avg
24-h avg
24-h avg
1-h max
8-h max
8-h max
8-h max
8-h max
8-h avg
8-h max
Mean
Concentration (ppbf
Summer:
1-h max: 44-117
8-h max: 30-99
Winter:
1-h max: 11-57
8-h max: 8-49
35.1-60
26.0
26.0d
All year: 26.8
May-September: 30.0
21 .4-48.7
U.S.: 13.3-38.4
Canada: 6.7-8.4
Europe:1 8.3-41 .9
16.1-58.8
20.0-62.8
41 .2-58.9
59.0-87.6
18.7-43.7
15.1-62.8
Upper Percentile
Concentrations (ppbf
Summer:
1-h max: 62-173
8-h max: 57-154
Winter:
1 -h max: 40-88
8-h max: 25-78
25th: 26.5-52
75th: 46.3-69
NR
NR
Maximum:
All year: 37.3
May-September: 47.2
NR
75th:
U.S.: 21.0-52.0
Canada: 8.7-12.5
Europe: 24.0-65.8
NR
75th: 27.2-74.8
75th: 47.0-71 .6
NR
75th: 38.4 -60.4
Max: 92.1 -131.8
Max: 34.3-1 46.2
75th: 19.8-75.9
6-223
-------
Study
Zanobetti and
Schwartz
(2008a)
Averaging
Location Years Time
1989-2000
(Winter: Dec-
Feb)
(Spring: Mar-
48 U.S. citiesc May) 8-h max
(Summer: June-
Aug)
(Autumn: Sept-
Nov)
Mean
Concentration (ppb)a
Winter: 16.5
Spring: 41.6
Summer: 47.8
Autumn: 33.5
Upper Percentile
Concentrations (ppbf
Max:
Winter: 40.6
Spring: 91.4
Summer: 103.0
Autumn: 91.2
aOzone concentrations were converted to ppb if the study presented them as ug/m3 by using the conversion factor of 0.51 assuming
standard temperature (25° C) and pressure (1 atm).
bStudy only reported median O3 concentrations.
°Cities with less than 75% observations in a season excluded. As a result, 29 cities examined in winter, 32 in spring, 33 in autumn,
and all 48 in the summer.
dBell et al. (2007) did not report mean O3 concentrations, however, it used a similar dataset as Bell et al. (2004) which consisted of
95 U.S. communities for 1987-2000. For comparison purposes the 24-h avg O3 concentrations for the 95 communities from Bell et
al. (2004) are reported here.
eStudy did not present air quality data for the summer months.
In addition to examining the relationship between short-term O3 exposure and all-
cause mortality, recent studies attempted to address the uncertainties that remained
upon the completion of the 2006 O3 AQCD. As a result, given the robust
associations between short-term O3 exposure and mortality presented across studies
in the 2006 O3 AQCD and supported in the new multicity studies (Figure 6-27). the
following sections primarily focus on the examination of previously identified
uncertainties in the O3-mortality relationship, specifically: confounding, effect
modification (i.e., sources of heterogeneity in risk estimates across cities), the
O3-mortality C-R relationship including lag structure (e.g., multiday effects and
mortality displacement), and O3 associations with cause-specific mortality. Focusing
specifically on these uncertainties allows for a more detailed characterization of the
relationship between short-term O3 exposure and mortality.
6.6.2.1 Confounding
Recent epidemiologic studies examined potential confounders of the O3-mortality
relationship. These studies specifically focused on whether PM and its constituents or
seasonal trends confounded the association between short-term O3 exposure and
mortality.
Confounding by PM and PM Constituents
An important question in the evaluation of the association between short-term O3
exposure and mortality is whether the relationship is confounded by particulate
matter, particularly the PM chemical components that are found in the "summer
haze" mixture which also contains O3. However, because of the temporal correlation
6-224
-------
among these PM components and O3, and their possible interactions, the
interpretation of results from copollutant models that attempt to disentangle the
health effects associated with each pollutant is challenging. Further complicating the
interpretation of copollutant results, at times, is the every-3rd or -6th day PM
sampling schedule employed in most locations, which limits the number of days
where both PM and O3 data is available.
The potential confounding effects of PM10 and PM2.5 on the O3-mortality
relationship were examined by Bell et al. (2007) using data on 98 U.S. urban
communities for the years 1987-2000 from the National Morbidity, Mortality, and
Air Pollution Study (NMMAPS). In this analysis the authors included PM as a
covariate in time-series models, and also examined O3-mortality associations on days
when O3 concentrations were below a specified value. This analysis was limited by
the small fraction of days when both PM and O3 data were available, due to the
every-3rd - or 6th -day sampling schedule for the PM indices, and the limited amount
of city-specific data for PM25 because it was only collected in most cities since 1999.
As a result, of the 91 communities with PM25 data, only 9.2% of days in the study
period had data for both O3 and PM2 5, resulting in the use of only 62 communities in
the PM2 5 analysis. An examination of the correlation between PM (PMi0 and PM25)
and O3 across various strata of daily PMi0 and PM25 concentrations found that
neither PM size fraction was highly correlated with daily O3 concentrations across
any of the strata examined. These results were also observed when using 8-h max
and 1-h max O3 exposure metrics. National and community-specific effect estimates
of the association between short-term O3 exposure and mortality were robust to
inclusion of PMi0 or PM25 in time-series models through the range of O3
concentrations (i.e., <10 ppb, 10-20, 20-40, 40-60, 60-80, and >80 ppb). Even with
the small number of days in which both PM2 5 and O3 data was available, the percent
increases in nonaccidental deaths per 10 ppb increase 24-h avg O3 concentrations at
lag 0-1 day were 0.22% (95% CI: -0.22, 0.65) without PM2.5 and 0.21% (95% CI:
-0.22, 0.64) with PM2.5 in 62 communities.
Although strong correlations between PM and O3 were not reported by Bell et al.
(2007) the patterns observed suggest regional differences in their correlation
(Table 6-44). Both PM10 and PM2 5 show positive correlations with O3 in the
Industrial Midwest, Northeast, Urban Midwest, and Southeast, especially in the
summer months, presumably, because of the summer peaking sulfate. However, the
mostly negative or weak correlations between PM and O3 in the summer in the
Southwest, Northwest, and southern California could be due to winter-peaking
nitrate. Thus, the potential confounding effect of PM on the O3-mortality relationship
could be influenced by the relative contribution of sulfate and nitrate, which varies
regionally and seasonally.
6-225
-------
Table 6-44 Correlations between PM and O3 by season and region.
No. of
Communities
Winter
Spring
Summer
Fall
Yearly
PM10
Industrial Midwest
Northeast
Urban Midwest
Southwest
Northwest
Southern California
Southeast
U.S.
19
15
6
9
11
7
25
93
0.37
0.34
0.24
0.00
-0.17
0.19
0.33
0.23
0.44
0.44
0.25
0.02
-0.20
0.08
0.35
0.26
0.44
0.36
0.22
-0.02
-0.13
0.12
0.31
0.24
0.39
0.44
0.26
0.10
-0.11
0.19
0.31
0.26
0.41
0.40
0.24
0.03
-0.16
0.14
0.32
0.25
PM2.5
Industrial Midwest
Northeast
Urban Midwest
Southwest
Northwest
Southern California
Southeast
U.S.
19
13
4
9
11
7
26
90
0.18
0.05
0.22
-0.15
-0.32
-0.25
0.38
0.09
0.39
0.26
0.31
-0.08
-0.34
-0.22
0.47
0.21
0.43
0.16
0.15
-0.17
-0.39
-0.25
0.30
0.12
0.44
0.43
0.32
-0.15
-0.24
-0.15
0.37
0.22
0.36
0.25
0.20
-0.14
-0.31
-0.23
0.39
0.16
Source: Bell et al. (2007).
6-226
-------
Raw Estimates
15-
10-
o
! 5-
.c
i
0 5
Without PM10
Posterior Estimates
10
1.5-
1.0-
3
I 0.5-
; 0.0 -
-0.5-
-1.0
-0.5
0.0 0.5
Without PM10
1.0
1.5
Note: The diagonal line indicates 1:1 ratio.
Source: Reprinted with permission of Informa UK Ltd, (Smith et al.. 2009b).
Figure 6-28 Scatter plots of Os mortality risk estimates with versus without
adjustment for PM™ in NMMAPS cities.
In an attempt to reassess a number of issues associated with the O3-mortality
relationship, including confounding, Smith et al. (2009b) re-analyzed the publicly
available NMMAPS database for the years 1987-2000. Similar to Bell et al. (2007).
the PMio data used in the Smith et al. (2009b) analysis consisted primarily of every -
6th day data. In analyses conducted to examine the potential confounding effects of
PM10, the authors reported that, in most cases, O3 mortality risk estimates were
reduced by between 22% and 33% in copollutant models. This is further highlighted
in Figure 6-28, which shows scatter plots of O3-mortality risk estimates with
adjustment for PMio versus without adjustment for PM10. Smith et al. (2009b) point
out that a larger fraction (89 out of 93) of the posterior estimates lie below the
diagonal line (i.e., estimates are smaller with PMio adjustment) compared to the raw
estimates (56 out of 93). This observation could be attributed to both sets of posterior
estimates being calculated by "shrinking towards the mean" along with the small
number of days where both PMio and O3 data was available. However, the most
prominent feature of these plots is that the variation of O3-mortality risk estimates
across cities is much larger than the impact of PMio adjustment on the O3-mortality
relationship.
6-227
-------
Franklin and Schwartz (2008) examined the sensitivity of O3 mortality risk estimates
to the inclusion of PM2.5 or PM chemical components associated with secondary
aerosols (e.g., sulfate [SO42~], organic carbon [OC], and nitrate [NO3-]) in
copollutant models. This analysis consisted of between 3 and 6 years of data from
May through September 2000-2005 from 18 U.S. communities. The association
between O3 and non-accidental mortality was examined in single-pollutant models
and after adjustment for PM2.5, sulfate, organic carbon, or nitrate concentrations.
The single-city effect estimates were combined into an overall estimate using a
random-effects model. In the single-pollutant model, the authors found a 0.89%
(95% CI: 0.45, 1.33%) increase in nonaccidental mortality with a 10 ppb increase in
same-day 24-hour summertime O3 concentrations across the 18 U.S. communities.
Adjustment for PM2.5 mass, which was available for 84% of the days, decreased the
O3-mortality risk estimate only slightly (from 0.88% to 0.79%), but the inclusion of
sulfate in the model reduced the risk estimate by 31% (from 0.85% to 0.58%).
However, sulfate data were only available for 18% of the days. Therefore, a
limitation of this study is the limited amount of data for PM2.5 chemical components
due to the every-3rd-day or every-6th-day sampling schedule. For example, when
using a subset of days when organic carbon measurements were available (i.e., 17%
of the available days), O3 mortality risk estimates were reduced to 0.51% (95% CI:
-0.36 to 1.36) in a single-pollutant model.
Consistent with the studies previously discussed, the results from Franklin and
Schwartz (2008) also demonstrate that the interpretation of the potential confounding
effects of copollutants on O3 mortality risk estimates is not straightforward as a result
of the PM sampling schedule employed in most cities. However, Franklin and
Schwartz (2008) find that O3-mortality risk estimates, although attenuated in some
cases (i.e., sulfate), remain positive. As presented in Figure 6-29. the regional and
city-to-city variations in O3 mortality risk estimates appear greater than the impact of
adjusting for copollutants. In addition, in some cases, a negative O3 mortality risk
estimate becomes even more negative with the inclusion of sulfate (e.g., Seattle) in a
copollutant model, or a null O3 mortality risk estimate becomes negative when
sulfate is included (e.g., Dallas and Detroit). Thus, the reduction in the overall O3
mortality risk estimate (i.e., across cities) needs to be assessed in the context of the
heterogeneity in the single-city estimates.
6-228
-------
Seattle
Riverside
Fresno
Sacramento
San Diego
El Paso
Dallas
Houston
Beaumont
Kansas City
St. Louis
Detroit
Cleveland
Pittsburgh
Buffalo
Rochester
Philadelphia
Boston
Source: Franklin and Schwartz (2008).
x
1
1 ' •
1 •
1
i n i
i —
i r~i
h^i
i •
i tn
• Ozone with sulfate 1
x Ozone alone 1
i ^ 1 .
i m 1
• ' i
i
• ' i
Pn1
• i
& 1
— i
1 ' • 1 '
• XI
1 n A i1
-5 0 5
Percent increase in mortality
with 10 ppb increase in ozone
Figure 6-29 Community-specific Os-mortality risk estimates for nonaccidental
mortality per 10 ppb increase in same-day 24-h average
summertime Os concentrations in single-pollutant models and
copollutant models with sulfate.
In the APHENA study, the investigators from the U.S. (NMMAPS), Canadian, and
European (APHEA2) multicity studies collaborated and conducted a joint analysis of
PMio and O3 using each of these datasets (Katsouyanni et al, 2009). For mortality,
each dataset consisted of a different number of cities and years of air quality data:
U.S. encompassed 90 cities with daily O3 data from 1987-1996 of which 36 cities
had summer only O3 measurements; Europe included 23 cities with 3-7 years of
daily O3 data during 1990-1997; and Canada consisted of 12 cities with daily O3 data
from 1987 to 1996. As discussed in Section 6.2.7.2, the APHENA study conducted
extensive sensitivity analyses, of which the 8 df/year results for both the penalized
spline (PS) and natural spline (NS) models are presented in the text for comparison
purposes, but only the NS results are presented in figures because alternative spline
models have previously been shown to result in similar effect estimates (FJEI, 2003).
6-229
-------
Additionally, for the Canadian results, figures contain risk estimates standardized to
a 40 ppb increment for 1-h max O3 concentrations, consistent with the rest of the
ISA, but also standardized to the approximate IQR across the Canadian cities as
discussed previously (Section 6.2.7.2).
In the three datasets, the authors found generally positive associations between short-
term O3 exposure and all-cause, cardiovascular, and respiratory mortality.
The estimated excess risks for O3 were larger for the Canadian cities than for the
U.S. and European cities. When examining the potential confounding effects of PMi0
on O3 mortality risk estimates, the sensitivity of the estimates varied across the data
sets and age groups. In the Canadian dataset, O3 risk estimates were modestly
reduced, but remained positive, when adjusting for PMi0 for all-cause mortality for
all ages in the PS (4.5% [95% CI: 2.2, 6.7%]) and NS (4.2% [95% CI: 1.9, 6.5%])
models to 3.8% (95% CI: -1.4, 9.8%) and 3.2% (95% CI: -2.2, 9.0%), respectively, at
lag 1 for a 40 ppb increase in 1-h max O3 concentrations (Figure 6-30 [and
Table 6-45]). However, adjusting for PM10 reduced O3 mortality risk estimates in the
> 75-year age group, but increased the risk estimates in the <75-year age group. For
cardiovascular and respiratory mortality more variable results were observed with O3
risk estimates being reduced and increased, respectively, in copollutant models with
PMio (Figure 6-30 [and Table 6-45]). Unlike the European and U.S. datasets, the
Canadian dataset only conducted copollutant analyses at lag 1; as a result, to provide
a comparison across study locations only the lag 1 results are presented for the
European and U.S. datasets in this section.
In the European data, O3 risk estimates were robust when adjusting for PMio in the
year-round data for all-cause, cardiovascular and respiratory mortality. When
restricting the analysis to the summer months moderate reductions were observed in
O3 risk estimates for all-cause mortality with more pronounced reductions in
respiratory mortality. In the U.S. data, adjusting for PMio moderately reduced O3
risk estimates for all-cause mortality in a year-round analysis at lag 1 (e.g., both the
PS and NS models were reduced from 0.18% to 0.13%) (Figure 6-30 [and
Table 6-45]). Similar to the European data, when restricting the analysis to the
summer months, in the United States. Ozone mortality risk estimates were
moderately reduced, but remained positive, when adjusting for PMio for all-cause
mortality. However, when examining cause-specific mortality risk estimates,
consistent with the results from the Canadian dataset, which employed a similar PM
sampling strategy (i.e., every-6th-day sampling), O3 risk estimates for cardiovascular
and respiratory mortality were more variable (i.e., reduced or increased in all-year
and summer analyses). Overall, the estimated O3 risks appeared to be moderately to
substantially sensitive to inclusion of PMio in copollutant models. Despite the
multicity approach, the mostly every-6th-day sampling schedule for PMio in the
Canadian and U.S. datasets greatly reduced the sample size and limits the
interpretation of these results.
6-230
-------
Location
APHENA-U.S.
APHENA-Canada
a
a
a
a
APHENA-Europe
Ages
All
>75
<75
>75
<75
All
All
>75
<75
All
All
>75
<75
>75
<75
All
All-Cause
Cardiovascular
Respiratory
All-Cause
Cardiovascular
Respiratory
All-Cause
Cardiovascular
Respiratory
-o-
All-Year
Summer
All-Year
Summer
All-Year
Summer
All-Year
All-Year
Summer
All-Year
Summer
All-Year
Summer
-10
10 15
% Increase
20
25
30
Note: Effect estimates are for a 40 ppb increase in 1-h max O3 concentrations at lag 1. All estimates are for the 8 df/year model with
natural splines. Circles represent all-year analysis results while diamonds represent summer season analysis results. Open circles
and diamonds represent copollutant models with PM10. Black = all-cause mortality; red = cardiovascular mortality; and
blue = respiratory mortality.
aRisk estimates from APHENA-Canada standardized to an approximate IQR of 5.1 ppb for a 1-h max increase in O3 concentrations
(see explanation in Section 6.2.7.2).
Figure 6-30 Percent increase in all-cause (nonaccidental) and cause-specific
mortality from natural spline models with 8 df/yr from the APHENA
study for single- and copollutant models.
6-231
-------
Table 6-45 Corresponding effect estimates for
Location* Mortality Ages Season
All P-II i«~o All
Figure 6-30.
Copollutant
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
% Increase (95% Cl)
1 .42 (0.08, 2.78)
1.02 (-1.40, 3.50)
4.31 (2.22, 6.45)
1 .90 (-0.78, 4.64)
1.10 (-1.33, 3.67)
0.47 (-4.61 , 5.79)
-0.1 6 (-3.02, 2.86)
1.34 (-3.63, 6.61)
3.58 (0.87, 6.37)
-1.1 7 (-6.1 8, 4.07)
3.18(0.31,6.12)
1 .26 (-4.46, 7.28)
2.46 (-1.87, 6.86)
3.50 (-4.23, 11.8)
6.04(1.18, 11.1)
7.03 (-3.48, 18.5)
4.15(1.90, 6.45)
PM10
PM10
0.52 (0.24, 0.80)a
3.18 (-2.18, 8.96)
0.40 (-0.28, 1.1 0)a
5.62(0.95, 10.7)
PM10
PM10
PM10
PM10
0.70(0.12, 1.30)a
1.90 (-9.03, 14.1)
0.24 (-1 .20, 1 .70)a
1.10 (-4.08, 6.61)
0.1 4 (-0.53, 0.82)a
-2.64 (-14.7, 11.5)
-0.34 (-2.00, 1 .40)a
0.87 (-6.40, 8.96)
PM10
PM10
0.11 (-0.84, 1.1 0)a
22.3 (-12.6, 71.3)
2.60 (-1.70, 7.1 0)a
6-232
-------
Location* Mortality Ages Season
All-year
All P-ll ir-c. All
Summer
> 75 All-year
<75 All-year
> 75 Summer
<75 Summer
All-year
Summer
Copollutant
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
% Increase (95% Cl)
1.02(0.39, 1.66)
1.26(0.47, 1.98)
2.06(1.10,2.94)
1.26(0.16,2.30)
1.10 (-0.47, 2.70)
1.1 8 (-0.55, 2.94)
1 .34 (-0.24, 2.94)
1.74 (-0.31, 3.75)
2.54 (0.39, 4.80)
1.58 (-0.70, 3.99)
1.66 (-0.70, 4.15)
1.66 (-1.02, 4.40)
1.42 (-1.02, 3.83)
1.42 (-1.02, 3.83)
4.31 (1.66, 7.11)
1.1 8 (-1.79, 4.31)
'Effect estimates from Figure 6-30.
aRisk estimates from APHENA-Canada standardized to an approximate IQR of 5.1 ppb for a 1-h max increase in O3 concentrations
(see explanation in Section 6.2.7.2).
Stafoggia et al. (2010) examined the potential confounding effects of PMi0 on the
Os-mortality relationship in individuals 35 years of age and older in 10 Italian cities
from 2001 to 2005. In a time-stratified case-crossover analysis, using data for the
summer months (i.e., April-September), the authors examined O3-mortality
associations across each city, and then obtained a pooled estimate through a
random-effects meta-analysis. Stafoggia et al. (2010) found a strong association with
nonaccidental mortality (9.2% [95% CI: 5.4, 13.0%] for a 30 ppb increase in 8-h max
O3 concentrations) in an unconstrained distributed lag model (lag 0-5) that persisted
in copollutant models with PM10 (9.2% [95% CI: 5.4, 13.7%]). Additionally, when
examining cause-specific mortality, the authors found positive associations between
short-term O3 exposure and cardiovascular (14.3% [95% CI: 6.7, 22.4%]),
cerebrovascular (8.5% [95% CI: 0.1, 16.3%]), and respiratory (17.6% [95% CI: 1.8,
35.6%]) mortality in single-pollutant models. In copollutant models, O3-mortality
effect estimates for cardiovascular and cerebrovascular mortality were robust to the
inclusion of PM10 (9.2% [95% CI: 5.4, 13.7%]) and 7.3% [95% CI: -1.2, 16.3%],
respectively), and attenuated, but remained positive, for respiratory mortality (9.2%
[95% CI: -6.9, 28.8%]). Of note, the correlations between O3 andPM10 across cities
were found to be generally low, ranging from (-0.03 to 0.49). The authors do not
specify the sampling strategy used for PMi0 in this analysis.
6-233
-------
Confounding by Seasonal Trend
The APHENA study (Katsouyanni et al., 2009), mentioned above, also conducted
extensive sensitivity analyses to identify the appropriate: (1) smoothing method and
basis functions to estimate smooth functions of time in city-specific models; and (2)
degrees of freedom to be used in the smooth functions of time, to adjust for seasonal
trends. Because O3 peaks in the summer and mortality peaks in the winter, not
adjusting or not sufficiently adjusting for the seasonal trend would result in an
apparent negative association between the O3 and mortality time-series. Katsouyanni
et al. (2009) examined the effect of the extent of smoothing for seasonal trends by
using models with 3 df/year, 8 df/year (the choice for their main model), 12 df/year,
and df/year selected using the sum of absolute values of partial autocorrelation
function of the model residuals (PACF) (i.e., choosing the degrees of freedom that
minimizes positive and negative autocorrelations in the residuals). Table 6-46
presents the results of the degrees of freedom analysis using alternative methods to
calculate a combined estimate: the Berkey et al. (1998) meta-regression and the two-
level normal independent sampling estimation (TLNISE) hierarchical method.
The results show that the methods used to combine single-city estimates did not
influence the overall results, and that neither 3 df/year nor choosing the df/year by
minimizing the sum of absolute values of PACF of regression residuals was
sufficient to adjust for the seasonal negative relationship between O3 and mortality.
However, it should be noted, the majority of studies in the literature that examined
the mortality effects of short-term O3 exposure, particularly the multicity studies,
used 7 or 8 df/year to adjust for seasonal trends, and in both methods a positive
association was observed between O3 exposure and mortality.
Table 6-46 Sensitivity of O3 risk estimates per 10 |jg/m3 increase in
24-h average O3 concentrations at lag 0-1 to alternative methods
for adjustment of seasonal trend, for all-cause mortality using
Berkey MLE and TLNISE Hierarchical Models.
Seasonality Control
3 df/year
8 df/year
12 df/year
PACF
Berkey
-0.54 (-0.88, 0.20)
0.30(0.11, 0.50)
0.34(0.15,0.53)
-0.62 (-1.01, -0.22)
TLNISE
-0.55 (-0.88, -0.22)
0.31 (0.09, 0.52)
0.33(0.12,0.54)
-0.62 (-0.98, -0.27)
Source: Reprinted with permission of Health Effects Institute (Katsouyanni et al., 2009).
6.6.2.2 Effect Modification
Epidemiologic studies have examined potential effect modifiers of the O3-mortality
relationship through the use of either: (1) time-invariant factors or (2) time-variant
factors. There have been several multicity studies that examined potential effect
6-234
-------
modifiers, or time-invariant factors, which may modify O3 mortality risk estimates.
These effect modifiers can be categorized into either individual-level or community-
level characteristics, which are traditionally examined in second stage regression
models. The results from these analyses also inform upon whether certain
populations are at greater risk of an O3-related health effect (Chapter Ł). In addition
to potentially modifying the association between short-term O3 exposure and
mortality, both individual-level and community-level characteristics may contribute
to the geographic pattern of spatial heterogeneity in O3 mortality risk estimates. As a
result, the geographic pattern of O3 mortality risk estimates is also evaluated in this
section. Although less common, this section also evaluates those studies that examine
effect modification by using time-variant factors, such as temperature and
copollutants that are included in first stage time-series regression models.
Time-Invariant Factors
Individual-Level Characteristics
Medina-Ramon and Schwartz (2008) conducted a case-only study in 48 U.S. cities to
identify populations potentially at increased risk to O3-related mortality for the
period 1989-2000 (May through September of each year [i.e., warm season]). A case-
only design predicts the occurrence of time-invariant characteristics among cases as a
function of the exposure level (Armstrong. 2003). For each potential effect modifier
(time-invariant individual-level characteristics), city-specific logistic regression
models were fitted, and the estimates were pooled across all cities. Furthermore, the
authors examined potential differences in individual effect modifiers according to
several city characteristics (e.g., mean O3 level, mean temperature, households with
central air conditioning, and population density) in a meta-regression. Across cities,
the authors found a 1.96% (95% CI: 1.14-2.82%) increase in mortality at lag 0-2 for
a 30 ppb increase in 8-h max O3 concentrations. Additionally, Medina-Ramon and
Schwartz (2008) examined a number of individual-level characteristics (e.g., age,
race) and chronic conditions (e.g., secondary causes of death) as effect modifiers of
the association between short-term O3 exposure and mortality. The authors found
that older adults (i.e., > 65), women >60 years of age, black race, and secondary
atrial fibrillation showed the greatest additional percent change in O3-related
mortality (Table 6-47). When examining city-level characteristics, the authors found
that older adults, black race, and secondary atrial fibrillation had a larger effect on O3
mortality risk estimates in cities with lower mean O3 concentrations. Of note, a
similar case-only study (Schwartz, 2005b) examined potential effect modifiers of the
association between temperature and mortality, which would be expected to find
results consistent with the Medina-Ramon and Schwartz (2008) study due to the high
correlation between temperature and O3. However, when stratifying days by
temperature Schwartz (2005b) found strong evidence that diabetes modified the
temperature-mortality association on hot days, which was not as evident when
examining the O3-mortality association in Medina-Ramon and Schwartz (2008). This
difference could be due to the study design and populations included in both studies,
a multicity study including all ages (Medina-Ramon and Schwartz, 2008) compared
6-235
-------
to a single-city study of individuals > 65 years of age (Schwartz, 2005b). However,
when examining results stratified by race, nonwhites were found to have higher
mortality risks on both hot and cold days, which provide some support for the
additional risk found for black race in Medina-Ramon and Schwartz (2008).
Individual-level factors that may result in increased risk of O3-related mortality were
also examined by Stafoggia et al. (2010). As discussed above, using a time-stratified
case-crossover analysis, the authors found an association between short-term O3
exposure and nonaccidental mortality in an unconstrained distributed lag model in 10
Italian cities (9.2% [95% CI: 5.4, 13.0%; lag 0-5 for a 30 ppb increase in 8-h max O3
concentrations). Stafoggia et al. (2010) conducted additional analyses to examine
whether age, sex, income level, location of death, and underlying chronic conditions
increased the risk of O3-related mortality, but data were only available for nine of the
cities for these analyses. Of the individual-level factors examined, the authors found
the strongest evidence for increased risk of O3-related mortality in individuals >
85 years of age (22.4% [95% CI: 15.0, 30.2%]), women (13.7% [95% CI: 8.5,
19.7%]), and out-of-hospital deaths (13.0% [95% CI: 6.0, 20.4%]). When focusing
specifically on out-of hospital deaths and the subset of individuals with chronic
conditions, Stafoggia et al. (2010) found the strongest association for individuals
with diabetes, which is consistent with the potentially increased risk of diabetics on
hot days observed in Schwartz (2005b).
6-236
-------
Table 6-47 Additional percent change in O3-related mortality for individual-
level characteristics.
Percentage (95% Cl)
Socio-demographic characteristics
Age 65 yr or older
Women
Women <60 yr old"
Women > 60 yr oldb
Black race
Low education
1.10
0.58
-0.09
0.60
0.53
-0.29
0.44, 1 .77
0.18,0.98
-0.76, 0.58
0.25, 0.96
0.19, 0.87
-0.81,0.23
Chronic conditions (listed as secondary cause)
Respiratory system diseases
Asthma
COPD
1.35
0.01
-0.31,3.03
-0.49, 0.52
Circulatory system diseases
Atherosclerosis
Atherosclerotic CVD
Atherosclerotic heart disease
Congestive heart disease
Atrial fibrillation
Stroke
-0.72
0.74
-0.38
-0.04
1.66
0.17
-1.89,0.45
-0.86, 2.37
-1.70,0.96
-0.39, 0.30
0.03, 3.32
-0.28, 0.62
Other diseases
Diabetes
Inflammatory diseases
0.19
0.18
-0.46, 0.84
-1 .09, 1 .46
"These estimates represent the additional percent change in mortality for persons who had the characteristic being examined
compared to persons who did not have the characteristic, when the mean O3 level of the previous 3 days increased 10 ppb.
These values were not standardized because they do not represent the actual effect estimate for the characteristic being
evaluated, but instead, the difference between effect estimates for persons with versus without the condition.
bCompared with males in the same age group.
Source: Reprinted with permission of Lippincott Williams & Wilkins (Medina-Ramon and Schwartz. 2008).
Additionally, Cakmak et al. (2011) examined the effect of individual-level
characteristics that may modify the O3-mortality relationship in 7 Chilean cities. In a
time-series analysis using a constrained distributed lag of 0-6 days, Cakmak et al.
(2011) found evidence for larger O3 mortality effects in individuals >75 years of age
compared to younger ages, which is similar to Medina-Ramon and Schwartz (2008)
and Stafoggia et al. (2010). Unlike the studies discussed above O3-mortality risk
estimates were found to be slightly larger in males (3.71% [95% CI: 0.79, 6.66] for a
40 ppb increase in max 8-h avg O3 concentrations), but were not significantly
different than those observed for females (3.00% [95% CI: 0.43, 5.68]). The major
focus of Cakmak et al. (2011) is the examination of the influence of SES indicators
(i.e., educational attainment, income level, and employment status) on the O3-
6-237
-------
mortality relationship. The authors found the largest risk estimates in the lowest SES
categories for each of the indicators examined this includes: primary school not
completed when examining educational attainment; the lowest quartile of income
level; and unemployed individuals when comparing employment status.
Overall, uncertainties exist in the interpretation of the potential effect modifiers
identified in Medina-Ramon and Schwartz (2008), Stafoggia et al. (2010), and
Cakmak et al. (2011) of the O3-mortality relationship due to the heterogeneity in O3-
mortality risk estimates across cities as highlighted in Smith et al. (2009b)
(Figure 6-28) and Franklin and Schwartz (2008) (Figure 6-29). In addition, it is likely
that individual-level factors identified in Medina-Ramon and Schwartz (2008),
Stafoggia et al. (2010), and Cakmak et al. (2011) only modify the O3-mortality
relationship and do not entirely explain the observed regional heterogeneity in
O3-mortality risk estimates.
Community-level Characteristics
Several studies also examined city-level (i.e., ecological) variables in an attempt to
explain the observed city-to-city variation in estimated O3-mortality risk estimates.
Bell and Dominici (2008) investigated whether community-level characteristics, such
as race, income, education, urbanization, transportation use, PM and O3
concentrations, number of O3 monitors, weather, and air conditioning use could
explain the heterogeneity in O3-mortality risk estimates across cities. The authors
analyzed 98 U.S. urban communities from NMMAPS for the period 1987-2000.
In the all-year regression model that included no community-level variables, a
20 ppb increase in 24-h avg O3 concentrations during the previous week was
associated with a 1.04% (95% CI: 0.56, 1.55) increase in mortality. Bell and
Dominici (2008) found that higher O3-mortality effect estimates were associated
with an increase in: percent unemployment, fraction of the population Black/African-
American, percent of the population that take public transportation to work; and with
a reduction in: temperatures and percent of households with central air conditioning
(Figure 6-31). The modification of O3-mortality risk estimates reported for city-
specific temperature and prevalence of central air conditioning in this analysis
confirm the result from the meta-analyses reviewed in the 2006 O3 AQCD.
The APHENA project (Katsouyanni et al., 2009) examined potential effect
modification of O3 risk estimates in the Canadian, European, and U.S. data sets using
a consistent set of city-specific variables. Table 6-48 presents the results from all age
analyses for all-cause mortality using all-year O3 data for the average of lag 0-1 day.
While there are several significant effect modifiers in the U.S. data, the results are
mostly inconsistent with the results from the Canadian and European data sets.
The positive effect modification by percentage unemployed and the negative effect
modification by mean temperature (i.e., a surrogate for air conditioning rate) are
consistent with the results reported by Bell and Dominici (2008) discussed above.
However, the lack of consistency across the data sets, even between the Canadian
and U.S. data, makes it difficult to interpret the results. Some of these associations
6-238
-------
may be due to coincidental correlations with other unmeasured factors that vary
regionally (e.g., mean SO2 tend to be higher in the eastern U.S.).
3 4 5 6 7 B
Percentage of populatron unemployed
0 10 20 30 40 50 60
Percentage of population
Black/Afncan American
4 -
n
50 55 60 65 70 75
Long-term temperature (°FJ
0 10 20 30 40 50
PQrcenlage of population taking
public transportation to wortt
ft'
t
0 20 W 60 80
Percentage of households with central AC
Note: The size of each circle corresponds to the inverse of the standard error of the community's maximum likelihood estimate. Risk
estimates are for a 10 ppb increase in 24-h avg O3 concentrations during the previous week.
Source: Reprinted with permission of Johns Hopkins Bloomberg School of Public Health (Bell and Dominici. 2008).
Figure 6-31 Ozone mortality risk estimates and community-specific
characteristics, U.S., 1987-2000.
6-239
-------
Table 6-48 Percent change in all-cause mortality, for all ages, associated with
a 40ppb increase in 1-h max O3 concentrations at Lag 0-1 at the
25th and 75th percentile of the center-specific distribution of
selected effect modifiers.
Canada
Effect
Modifier
NO2 CV
Mean SO2
03CV
Mean
NO2/PM10
Mean
Temperature
% > 75 yr
Age-
standardized
Mortality
%
Unemployed
25th
Percentile
Estimate
(95% Cl)
3.10
(1 .90, 4.40)
2.22
(0.71 , 3.83)
2.86
(0.79, 5.05)
3.91
(2.54, 5.29)
2.86
(0.95, 4.72)
2.22
(0.79, 3.58)
2.62
(0.79, 4.48)
2.78
(1 .42, 4.07)
75th
Percentile
Estimate
(95% Cl)
3.99
(2.38, 5.62)
4.72
(2.94,6.61)
3.50
(2.14, 4.89)
2.54
(0.95, 4.15)
3.50
(2.22, 4.89)
4.23
(3.02, 5.54)
4.07
(2.22, 5.87)
3.75
(2.54, 4.89)
t-
value
1.33
2.16
0.60
-1.58
0.83
2.68
1.14
1.88
25th
Percentile
Estimate
(95% Cl)
1.66
(0.71, 2.62)
1.58
(0.47, 2.62)
2.62
(1 .50, 3.75)
1.74
(0.87, 2.70)
1.58
(0.39, 2.86)
1.50
(0.55, 2.46)
1.10
(-0.16,2.38)
1.42
(-0.47, 3.34)
Europe
75th
Percentile
Estimate
(95% Cl)
1.34
(-0.08, 2.78)
1.66
(0.39, 2.86)
1.10
(0.24, 1 .98)
1.50
(0.47, 2.62)
1.58
(0.31, 2.78)
1.82
(0.55, 3.10)
1.98
(0.79, 3.26)
1.34
(-0.47,3.18)
t-
value
-0.49
0.16
-2.65
-0.43
-0.04
0.52
1.07
-0.07
25th
Percentile
Estimate
(95% Cl)
1.26
(0.47, 1 .98)
0.47
(-0.47, 1 .42)
0.16
(-0.70, 1.10)
-0.08
(-1.02,0.95)
2.14
(1 .34, 2.94)
1.02
(0.24, 1.90)
0.00
(-0.94, 0.87)
0.16
(-0.78,1.18)
U.S.
75th
Percentile
Estimate
(95% Cl)
0.08
(-0.78, 0.95)
1.98
(1.10, 2.94)
1.50
(0.71, 2.22)
1.26
(0.47, 2.06)
0.00
(-0.78, 0.79)
1.02
(0.31, 1.74)
1.58
(0.87, 2.38)
1.50
(0.71,2.30)
t-
value
-2.87
2.79
2.68
2.64
-4.40
-0.02
3.81
2.45
Source: Adapted with permission of Health Effects Institute; Katsouvanni et al. (2009).
Regional Pattern of Ozone-Mortality Risk Estimates
In addition to examining whether individual- and community-level factors modify
the O3-mortality association, studies have also examined whether these associations
varied regionally within the United States. Bell and Dominici (2008), in the study
discussed above, also noted that O3-mortality risk estimates were higher in the
Northeast (1.44% [95% Cl: 0.78, 2.10%]) and Industrial Midwest (0.73% [95% Cl:
0.11, 1.35%]), while null associations were observed in the Southwest and Urban
Midwest (Table 6-49). The regional heterogeneity in O3-mortality risk estimates was
further reflected by Bell and Dominici (2008) in a map of community-specific
Bayesian O3-mortality risk estimates (Figure 6-32). It is worth noting that in the
analysis of PMi0 using the same data set, Peng et al. (2005) also found that both the
Northeast and Industrial Midwest showed particularly elevated effects, especially
during the summer months. As mentioned above, although no evidence for
confounding of O3 mortality risk estimates by PMi0 was observed, Bell et al. (2007)
did find regional differences in the correlation between O3 and PMi0. Thus, the
heterogeneity in O3 mortality risk estimates may need to be examined as a function
of the correlation between PM and O3.
6-240
-------
Smith et al. (2009b), as discussed earlier, also examined the regional difference in O3
mortality risk estimates across the same seven regions and similarly found evidence
for regional heterogeneity. In addition, Smith et al. (2009b) constructed spatial maps
of the risk estimates by an extension of a hierarchical model that allows for spatial
auto-correlation among the city-specific random effects. Figure 6-33 presents the
spatial map of O3 mortality coefficients from the Smith et al. (2009b) analysis that
used 8-h max O3 concentrations during the summer. The results from the Bell and
Dominici (2008) analysis (Figure 6-32) shows much stronger apparent heterogeneity
in O3-mortality risk estimates across cities than the smoothed map from Smith et al.
(2009b) (Figure 6-33). but both maps generally show larger risk estimates in the
eastern region of the U.S.
6-241
-------
Table 6-49 Percentage increase in daily mortality for a 10 ppb increase in
24-h average O$ concentrations during the previous week by
geographic region in the U.S., 1987-2000.
No. of Communities
Regional Estimate
95% PI*
Regional results
Industrial Midwest
Northeast
Northwest
Southern California
Southeast
Southwest
Urban Midwest
20
16
12
7
26
9
7
0.73
1.44
0.08
0.21
0.38
-0.06
-0.05
0.11, 1.35
0.78, 2.10
-0.92, 1.09
-0.46, 0.88
-0.07, 0.85
-0.92, 0.81
-1.28, 1.19
National results
All continental communities
All communities
*PI, posterior interval.
Source: Reprinted with permission of Johns
97
98
Hopkins Bloomberg School
0.51
0.52
of Public Health (Bell and Dominici, 2008).
0.27, 076
0.28, 0.77
<0.0
Source: Reprinted with permission of Johns Hopkins Bloomberg School of Public Health, (Bell and Dominici. 2008).
Figure 6-32 Community-specific Bayesian O3-mortality risk estimates in 98
U.S. communities.
6-242
-------
8H: summer
Source: Reprinted with permission of Informa UK Ltd. (Smith et al.. 2009b).
Figure 6-33 Map of spatially dependent Os-mortality coefficients for 8-h max
concentrations using summer data.
Time-Variant Factors
To date, only a few time-series studies have investigated the potential interaction
between O3 exposure and copollutants or weather variables in first stage regression
models. This can be attributed to the moderate to high correlation between O3 and
these covariates, which makes such investigations methodologically challenging.
Ren et al. (2008) examined the possible synergistic effect between O3 and
temperature on mortality in the 60 largest eastern U.S. communities from the
NMMAPS data during the warm months (i.e., April to October) from 1987-2000.
This analysis was restricted to the eastern areas of the U.S. (i.e., Northeast, Industrial
Midwest and Southeast) because a previous study which focused specifically on the
eastern U.S. found that temperature-mortality patterns differ between the northeast
and southeast regions possibly due to climatic differences (Curriero et al.. 2002).
To examine possible geographic differences in the interaction between temperature
and O3, Ren et al. (2008) further divided the NMMAPS regions into the Northeast,
which included the Northeast and Industrial Midwest regions (34 cities), and the
Southeast, which included the Southeast region (26 cities). The potential synergistic
effects between O3 and temperature were examined using two different models.
Model 1 included an interaction term in a Generalized Additive Model (GAM) for
O3 and maximum temperature (3-day avg values were used for both terms) to
examine the bivariate response surface and the pattern of interaction between the two
variables in each community. Model 2 consisted of a Generalized Linear Model
(GLM) that used interaction terms to stratify by "low," "moderate," and "high"
6-243
-------
temperature days using the first and third quartiles of temperature as cut-offs to
examine the percent increase in mortality in each community. Furthermore, a two-
stage Bayesian hierarchical model was used to estimate the overall percent increase
in all-cause mortality associated with short-term O3 exposure across temperature
levels and each region using model 2. The same covariates were used in both model
1 and 2. The bivariate response surfaces from model 1 suggest possible interactive
effects between O3 and temperature although the interpretation of these results is not
straightforward due to the high correlation between these terms. The apparent
interaction between temperature and O3 as evaluated in model 2 varied across
geographic regions. In the northeast region, a 20 ppb increase in 24-h avg O3
concentrations at lag 0-2 was associated with an increase of 4.49% (95% posterior
interval [PI]: 2.39, 6.36%), 6.21% (95% PI: 4.47, 7.66%) and 12.8% (95% PI: 9.77,
15.7%) in mortality at low, moderate and high temperature levels, respectively.
The corresponding percent increases in mortality in the southeast region were 2.27%
(95% PI: -2.23, 6.46%) for low temperature, 3.02% (95% PI: 0.44, 5.70%) for
moderate temperature, and 2.60% (95% PI: -0.66, 6.01%) for high temperature.
When examining the relationship between temperature and O3-related mortality, the
results reported by Ren et al. (2008) (i.e., higher O3-mortality risks on days with
higher temperatures) may appear to contradict the results of Bell and Dominici
(2008) described earlier (i.e., communities with higher temperature have lower
O3-mortality risk estimates). However, the observed difference in results can be
attributed to the interpretation of effect modification in a second-stage regression
which uses long-term average temperatures, as was performed by Bell and Dominici
(2008), compared to a first-stage regression that examines the interaction between
daily temperature and O3-related mortality. In this case, the second-stage regression
results from Bell and Dominici (2008) indicate that a city with lower temperatures,
on average, tend to show a stronger O3 mortality effect, whereas, in the first-stage
regression performed by Ren et al. (2008). the days with higher temperature tend to
show a larger O3-mortality effect. This observed difference may in part reflect the
higher air conditioning use in communities with higher long-term average
temperatures. Therefore, the findings from Ren et al. (2008) indicating generally
lower O3 risk estimates in the southeast region where the average temperature is
higher than in the northeast region is consistent with the regional results reported by
Bell and Dominici (2008). As demonstrated by the results from both Ren et al.
(2008) and Bell and Dominici (2008) caution is required when interpreting results
from studies that examined interactive effects using two different approaches because
potential effect modification as suggested in a second-stage regression generally does
not provide evidence for a short-term interaction examined in a first-stage regression.
Overall, further examination of the potential interactive (synergistic) effects of O3
and covariates in time-series regression models is required to more clearly
understand the factors that may influence O3 mortality risk estimates.
6-244
-------
6.6.2.3 Evaluation of the O3-Mortality C-R Relationship and
Related Issues
Evaluation of the O3-mortality C-R relationship is not straightforward because the
evidence from multicity studies (using log-linear models) suggests that O3-mortality
associations are highly heterogeneous across regions. In addition, there are numerous
issues that may influence the shape of the O3-mortality C-R relationship and the
observed association between short-term O3 exposure and mortality that warrant
examination including: multi-day effects (distributed lags), mortality displacement
(i.e., hastening of death by a short period), potential adaptation, and the exposure
metric used to compute risks (e.g., 1-hour daily max versus 24-h avg). The following
section presents the recent studies identified that conducted an initial examination of
these issues.
Multiday Effects, Mortality Displacement, and Adaptation
The pattern of positive lagged associations followed by negative associations in a
distributed lag model may be considered an indication of "mortality displacement"
(i.e., deaths are occurring in frail individuals and exposure is only moving the day of
death to a day slightly earlier). Zanobetti and Schwartz (2008b) examined this issue
in 48 U.S. cities during the warm season (i.e., June-August) for the years 1989-2000.
In an initial analysis, the authors applied a GLM to examine same-day O3-mortality
effects, and in the model included an unconstrained distributed lag for apparent
temperature to take into account the effect of temperature on the day death occurred
and the previous 7 days. To examine mortality displacement Zanobetti and Schwartz
(2008b) refit models using two approaches: an unconstrained and a smooth
distributed lag each with 21-day lags for O3. In this study, all-cause mortality as well
as cause-specific mortality (i.e., cardiovascular, respiratory, and stroke) were
examined for evidence of mortality displacement. The authors found a 0.96%
(95% CI: 0.60, 1.30%) increase in all-cause mortality across all 48 cities for a 30 ppb
increase in 8-h max O3 concentrations at lag 0 whereas the combined estimate of the
unconstrained distributed lag model (lag 0-20) was 1.54% (95% CI: 0.15, 2.91%).
Similarly, when examining the cause-specific mortality results (Table 6-50). larger
risk estimates were observed for the distributed lag model compared to the lag 0 day
estimates. However, for stroke a slightly larger effect was observed at lags 4-20
compared to lags 0-3 suggesting a larger window for O3-induced stroke mortality.
This is further supported by the sum of lags 0 through 20 days showing the greatest
effect. Overall, these results suggest that estimating the mortality risk using a
single day of O3 exposure may underestimate the public health impact, but the extent
of multi-day effects appear to be limited to a few days. This is further supported by
the shape of the combined smooth distributed lag (Figure 6-34). It should be noted
that the proportion of total variation in the effect estimates due to the between-cities
heterogeneity, as measured by I2 statistic, was relatively low (4% for the lag 0
estimates and 21% for the distributed lag), but 21 out of the 48 cities exhibited null
or negative estimates. As a result, the estimated shape of the distributed lag cannot be
6-245
-------
interpreted as a general form of lag structure of associations applicable to all the
cities included in this analysis.
Samoli et al. (2009) also investigated the temporal pattern of mortality effects in
response to short-term exposure to O3 in 21 European cities that were included in the
APHEA2 project. Using a method similar to Zanobetti and Schwartz (2008b). the
authors applied unconstrained distributed lag models with lags up to 21 days in each
city during the summer months (i.e., June through August) to examine the effect of
O3 on all-cause, cardiovascular, and respiratory mortality. They also applied a
generalized additive distributed lag model to obtain smoothed distributed lag
coefficients. However, unlike Zanobetti and Schwartz (2008b), Samoli et al. (2009)
controlled for temperature using a linear term for humidity and an unconstrained
distributed lag model of temperature at lags 0-3 days. The choice of 0- through 3-day
lags of temperature was based on a previous European multicity study (Baccini et al.,
2008), which suggested that summer temperature effects last only a few days. Upon
combining the individual city estimates across cities in a second stage regression,
Samoli et al. (2009) found that the estimated effects on respiratory mortality were
extended for a period of two weeks. However, for all-cause and cardiovascular
mortality, the 21-day distributed lag models yielded null or (non-significant) negative
estimates (Table 6-51). Figure 6-35 shows the distributed lag coefficients for all-
cause mortality, which exhibit a declining trend and negative coefficients beyond
5-day lags. The authors' interpretation of these results was that "using single-day
exposures may have overestimated the effects on all-cause and cardiovascular
mortality, but underestimated the effects on respiratory mortality." Thus, the results
in part suggest evidence of mortality displacement for all-cause and cardiovascular
mortality.
6-246
-------
Table 6-50 Estimated effect of a 10 ppb increase in 8-h max O3 concentrations
on mortality during the summer months for single-day and
distributed lag models.
% (Percentage)
95% Cl
Total mortality
LagO
Sum lags 0-20
Sum lags 0-3
Sum lags 4-20
0.32
0.51
0.53
-0.02
0.20, 0.43
0.05, 0.96
0.28, 0.77
-0.35, 0.31
Cardiovascular mortality
LagO
Sum lags 0-20
Sum lags 0-3
Sum lags 4-20
0.47
0.49
0.80
-0.23
0.30, 0.64
-0.01 , 1 .00
0.48, 1.13
-0.67, 0.22
Respiratory mortality
LagO
Sum lags 0-20
Sum lags 0-3
Sum lags 4-20
0.54
0.61
0.83
-0.24
0.26, 0.81
-0.41 , 1 .65
0.38, 1 .28
-1.08,0.60
Stroke
LagO
Sum lags 0-20
Sum lags 0-3
Sum lags 4-20
0.37
2.20
0.92
1.26
0.01,0.74
0.76, 3.67
0.26, 1 .59
0.05, 2.49
Source: Reprinted with permission of American Thoracic Society, Zanobetti and Schwartz (2008b).
6-247
-------
OJ
o
.g
Q. T .
O. O
p
o '
a
pi
OJ
9'
10
15
20
Day Lag
Source: Reprinted with permission of American Thoracic Society (Zanobetti and Schwartz. 2008b).
Note: The triangles represent the percent increase in all-cause mortality for a 10 ppb increase in 8-h max O3 concentrations at each
lag while the shaded areas are the 95% point-wise confidence intervals.
Figure 6-34 Estimated combined smooth distributed lag for 48 U.S. cities
during the summer months.
6-248
-------
Table 6-51 Estimated percent increase in cause-specific mortality (and
95% CIs) for a 10-ug/m3 increase in 8-h daily max O3 during
June-August.
Fixed effects % (95% Cl) Random effects % (95% Cl)
Total mortality3
LagO
Average lags 0-1
Sum lags 0-20, unconstrained
Sum lags 0-20, penalized
0.28(0.11,0.45)
0.24(0.15,0.34)
0.01 (-0.40,0.41)
0.01 (-0.41,0.42)
0.28 (0.07, 0.48)
0.22 (0.08, 0.35)
-0.54 (-1 .28, 0.20)
-0.56 (-1.30, 0.19)
Cardiovascular mortality3
LagO
Average lags 0-1
Sum lags 0-20, unconstrained
Sum lags 0-20, penalized
0.43(0.18,0.69)
0.33(0.19,0.48)
-0.33 (-0.93, 0.29)
-0.32 (-0.92, 0.28)
0.37 (0.05, 0.69)
0.25 (0.03, 0.47)
-0.62 (-1 .47, 0.24)
-0.57 (-1 .39, 0.26)
Respiratory mortality3
LagO
Average lags 0-1
Sum lags 0-20, unconstrained
Sum lags 0-20, penalized
0.36 (-0.21, 0.94)
0.40(0.11,0.70)
3.35(1.90,4.83)
3.66 (2.25, 5.08)
0.36 (-0.21, 0.94)
0.40(0.11,0.70)
3.35(1.90,4.83)
3.66 (2.25, 5.08)
"Analysis for the same day (lag 0), the average of the same and previous day (lag 0-1), the unconstrained distributed lag model for
the sum of 0-20 days and the penalized distributed lag model (lag 0-20)
Source: Used with permission of BMJ Group (Samoli et al.. 2009).
Although the APHENA project (Katsouyanni et al.. 2009) did not specifically
investigate mortality displacement and therefore did not consider longer lags
(e.g., lag >3 days), the study did present O3 risk estimates for lag 0-1, lag 1, and a
distributed lag model of 0-2 days in the Canadian, European, and U.S. datasets.
Katsouyanni et al. (2009) found that the results vary somewhat across the regions,
but, in general, there was no indication that the distributed lag model with up to a
2-day lag yielded meaningfully larger O3 mortality risk estimates than the lag 0-1
and lag 1 results. For example, for all-cause mortality, using the model with natural
splines and 8 df/year to adjust for seasonal trends, the reported percent excess risk for
mortality for a 40 ppb increase in 1-h max O3 concentrations for lag 0-1, lag 1, and
the distributed lag model (lag 0-2) was 2.70% (95% Cl: 1.02, 4.40%), 1.42%
(95% Cl: 0.08, 2.78%), and 3.02% (95% Cl: 1.10, 4.89%), respectively. Thus, the
observed associations appear to occur over a short time period, (i.e., a few days).
Similarly, the Public Health and Air Pollution in Asia (PAPA) study (Wong et al.,
2010) also examined multiple lag days (i.e., lag 0, lag 0-1, and lag 0-4), and although
it did not specifically examine mortality displacement it does provide additional
evidence regarding the timing of mortality effects proceeding O3 exposure. In a
combined analysis using data from all four cities examined (Bangkok, Hong Kong,
Shanghai, and Wuhan), excess risk estimates at lag 0-4 were larger than those at lag
6-249
-------
0 or lag 0-1 in both fixed and random effect models (results not presented
quantitatively). The larger risk estimates at lag 0-4 can primarily be attributed to the
strong associations observed in Bangkok and Shanghai. However, it is worth noting
that Bangkok differs from the three Chinese cities included in this analysis in that it
has a tropical climate and does not exhibit seasonal patterns of mortality. As a result,
Wong et al. (2010) examined the O3-mortality associations at lag 0-1 in only the
three Chinese cities and found that risk estimates were slightly reduced from 2.26%
(95% CI: 1.36, 3.16) in the 4 city analysis to 1.84% (0.77, 2.86) in the 3 city analysis
for a 30 ppb increase in 8-h max O3 concentrations. Overall, the PAPA study further
supports the observation of the APHENA study that associations between O3 and
mortality occur over a relatively short-time period, but also indicates that it may be
difficult to interpret O3-mortality associations across cities with different climates
and mortality patterns.
When comparing the studies that explicitly examined the potential for mortality
displacement in the O3-mortality relationship, the results from Samoli et al. (2009),
which provide evidence that suggests mortality displacement, are not consistent with
those reported by Zanobetti and Schwartz (2008b). However, the shapes of the
estimated smooth distributed lag associations are similar (Figure 6-34 and
Figure 6-35). A closer examination of these figures shows that in the European data
beyond a lag of 5 days the estimates remain negative whereas in the U.S. data the
results remain near zero for the corresponding lags. These observed difference could
be due to the differences in the model specification between the two studies,
specifically the use of: an unconstrained distributed lag model for apparent
temperature up to 7 previous days (Zanobetti and Schwartz, 2008b) versus a linear
term for humidity and an unconstrained distributed lag model of temperature up to 3
previous days (Samoli et al.. 2009): and natural cubic splines with 2 df per season
(Zanobetti and Schwartz. 2008b) versus dummy variables per month per year to
adjust for season (Samoli et al.. 2009). It is important to note that these differences in
model specification may have also influenced the city-to-city variation in risk
estimates observed in these two studies (i.e., homogenous estimates across cities in
Zanobetti and Schwartz (2008b) and heterogeneous estimates across cities in Samoli
et al. (2009). Overall, the evidence of mortality displacement remains unclear, but
Samoli et al. (2009). Zanobetti and Schwartz (2008b). and Katsouvanni et al. (2009)
all suggest that the positive associations between O3 and mortality are observed
mainly in the first few days after exposure.
6-250
-------
Ł
Note: The triangles represent the percent increase in all-cause mortality for a 10 ug/m3 increase in 8-h max O3 concentrations at
each lag; the shaded area represents the 95% CIs.
Source: Reprinted with permission of BMJ Group (Samoli et al.. 2009).
Figure 6-35 Estimated combined smooth distributed lag in 21 European cities
during the summer (June-August) months.
Adaptation
Controlled human exposure studies have demonstrated an adaptive response to O3
exposure for respiratory effects, such as lung function decrements, but this issue has
not been examined in the epidemiologic investigation of mortality effects of O3.
Zanobetti and Schwartz (2008a) examined if there was evidence of an adaptive
response in the O3-mortality relationship in 48 U.S. cities from 1989 to 2000 (i.e., the
same data analyzed in Zanobetti and Schwartz (2008b). The authors examined all-
cause mortality using a case-crossover design to estimate the same-day (i.e., lag 0)
effect of O3, matched on referent days from every-3rd-day in the same month and
year as the case. Zanobetti and Schwartz (2008a) examined O3-mortality associations
by: season, month in the summer season (i.e., May through September), and age
categories in the summer season (Table 6-52). The estimated O3 mortality risk
estimate at lag 0 was found to be highest in the summer (1.51% [95% CI: 1.14,
1.87%]; lag 0 for a 30 ppb increase in 8-h max O3 concentrations), and, within the
warm months, the association was highest in My (1.96% [95% CI: 1.42, 2.48%];
6-251
-------
lag O).1 Upon further examination of the summer months, the authors also observed
diminished effects in August (0.84% [95% CI: 0.33, 1.39%]; lag 0). Based on these
results, the authors concluded that the mortality effects of O3 appear diminished later
in the O3 season.
Table 6-52 Percent excess all-cause mortality per 10 ppb increase in daily
8-h max O3 on the same day, by season, month, and age groups.
% 95% CI
By Season
Winter
-0.13
-0.56, 0.29
Spring
0.35
0.16, 0.54
Summer
0.50
0.38, 0.62
Fall
0.05
-0.14,0.24
By Month
May
0.48
0.28, 0.68
June
0.46
0.24, 0.68
July
0.65
0.47, 0.82
August
0.28
0.11,0.46
September
-0.09
-0.35, 0.16
By Age Group
0-20
0.08
-0.42, 0.57
21-30
0.10
-0.67, 0.87
31-40
0.07
-0.38, 0.52
41-50
0.08
-0.27, 0.43
51-60
0.54
0.19, 0.89
61-70
0.38
0.16,0.61
71-80
0.50
0.32, 0.67
80
0.29
0.13, 0.44
Source: Zanobetti and Schwartz (2008a).
To further evaluate the potential adaptive response observed in Zanobetti and
Schwartz (2008a) the distribution of the O3 concentrations across the 48 U.S. cities
during July and August was examined. Both July and August were found to have
comparable means of 48.6 and 47.9 ppb with a reported maximum value of 97.9 and
96.0 ppb, respectively. Thus, the observed reduction in O3-related mortality effect
estimates in August (0.84%) compared to July (1.96%) appears to support the
existence of an adaptive response. However, unlike an individual's adaptive response
1 These values have been standardized to the increment used throughout the ISA for max 8-h avg increase in O3 concentrations of
30 ppb. These values differ from those presented in Table 6-52: from Zanobetti and Schwartz (2008a) because the authors
presented values for a 10 ppb increase in max 8-h avg O3 concentrations.
6-252
-------
to decrements in lung function from short-term O3 exposure, an examination of
mortality prevents a direct observation of adaptation. Rather, for mortality the
adaptation hypothesis is tested with a tacit assumption that, whatever the mechanism
for O3-induced mortality, the risk of death from short-term O3 exposure is reduced
over the course of the summer months through repeated exposures. This idea would
translate to a smaller population that would die from O3 exposure toward the end of
summer. This may complicate the interpretation of the distributed lag coefficients
with long lag periods because the decreased coefficients may reflect diminished
effects of the late summer, rather than diminished effects that are constant across the
summer. These intertwined issues need to be investigated together in future research.
Exposure Metric
When examining the association between short-term O3 exposure and mortality it is
also important to consider the exposure metric used (i.e., 24-h avg, 8-h max, and
1-h max). To date, only a few studies have conducted analyses to examine the impact
of different exposure metrics on O3 mortality risk estimates. In Smith et al. (2009b),
the authors examined the effect of different exposure metrics (i.e., 24-h avg, 8-h max,
and 1-h max) on O3-mortality regression coefficients. When examining whether
there are differences in city-specific risk estimates when using different exposure
metrics, Smith et al. (2009b) found a rather high correlation (r = 0.7 - 0.8) between
risk estimates calculated using 24-h avg versus 8-h max and 1-h max versus 8-h max
averaging times. These results are consistent with the correlations reported by
Darrow et al. (201 la) (Section 6.2.7.3) between the 8-h max and 24- avg exposure
metrics.
In addition to these recent studies published since the 2006 O3 AQCD, Gryparis et al.
(2004) also supports the high correlation between 1-h max and 8-h max O3
concentrations reported in Smith et al. (2009b) and Darrow et al. (201 la) and the
subsequent high degree of similarity between mortality risk estimates calculated
using either metric. Although only a limited number of studies have examined the
effect of different exposure metrics on O3-mortality risk estimates, these studies
suggest relatively comparable results across the exposure metrics used.
Ozone-Mortality C-R Relationship and Threshold Analyses
Several of the recent studies evaluated have applied a variety of statistical approaches
to examine the shape of the O3-mortality C-R relationship and whether a threshold
exists. The approach used by Bell et al. (2006) consisted of applying four statistical
models to the NMMAPS data, which included 98 U.S. communities for the period
1987-2000. These models included: a linear analysis (i.e., any change in O3
concentration can be associated with mortality) (Model 1); a subset analysis
(i.e., examining O3-mortality relationship below a specific 24- avg concentration,
ranging from 5 to 60 ppb) (Model 2); a threshold analysis (i.e., assuming that an
association between O3 and mortality is observed above a specific concentration and
6-253
-------
not below it, using the threshold values set at an increment of 5 ppb between 0 to
60 ppb and evaluating a presence of a local minima in AICs computed at each
increment) (Model 3); and nonlinear models using natural cubic splines with
boundary knots placed at 0 and 80 ppb, and interior knots placed at 20 and 40 ppb
(Model 4). A two-stage Bayesian hierarchical model was used to examine these
models and O3-mortality risk estimates at the city-level in the first stage analysis and
aggregate estimates across cities in the 2nd stage analysis using the average of 0- and
1-day lagged 24-h avg O3 concentrations. The results from all of these models
suggest that if a threshold exists it does so well below the current O3 NAAQS. When
restricting the analysis to all days when the 1997 O3 NAAQS 8-hour standard
(i.e., 84 ppb daily 8-h max) is met in each community, Bell et al. (2006) found there
was still a 0.60% (95% PI: 0.30, 0.90%) increase in mortality per 20 ppb increase in
24-h avg O3 concentrations at lag 0-1. Figure 6-36 shows the combined C-R curve
obtained using the nonlinear model (Model 4). Although these results suggest the
lack of threshold in the O3-mortality relationship, it is difficult to interpret such a
curve because: (1) there is uncertainty around the shape of the C-R curve at 24-h avg
O3 concentrations generally below 20 ppb, and (2) the C-R curve does not take into
consideration the heterogeneity in O3-mortality risk estimates across cities.
6-254
-------
•= 5
S 4
o
= 3
i 2
o
Central estimate
95% posterior interval
0 20 40 GO 80
Average of same and previous days' 03 (ppb)
Source: Bell et al. (2006)
Figure 6-36 Estimated combined C-R curve for nonaccidental mortality and
24-hour average Os concentrations at lag 0-1 using the nonlinear
(spline) model.
Using the same NMMAPS dataset as Bell et al. (2006). Smith et al. (2009b) further
examined the O3-mortality C-R relationship. Similar to Bell et al. (2006), Smith et al.
(2009b) conduct a subset analysis, but instead of restricting the analysis to days with
O3 concentrations below a cutoff the authors only include days above a defined
cutoff in the analysis. The results of this "reversed subset" approach are in line with
those reported by Bell et al. (2006); consistent positive associations at all cutoff
points up to a defined concentration where the total number of days with 24-h avg O3
concentrations above a value are so limited that the variability around the central
estimate is increased. In the Smith et al. (2009b) analysis this observation was
initially observed at 45 ppb, with the largest variability at 60 ppb; however, unlike
Bell et al. (2006) where 73% of days are excluded when subsetting the data to less
than 20 ppb, the authors do not detail the number of days of data included in the
subset analyses at higher concentrations. In addition to the subset analysis, Smith et
al. (2009b) examined the shape of the C-R curve using a piecewise linear approach
with cutpoints at 8-h avg concentrations of 40 ppb, 60 ppb, and 80 ppb. Smith et al.
(2009b) found that the shape of the C-R curve is similar to that reported by Bell et al.
(2006) (Figure 6-36). but argue that slopes of the (3 for each piece of the curve are
highly variable with the largest variation in the 60-80 ppb range. However, the larger
6-255
-------
variability around the (3 between 60-80 would be expected due to the small number of
days with O3 concentrations within that range in an all-year analysis. This result is
consistent with that observed by Bell et al. (2006), which is presented in Figure 6-36.
The APHENA project (Katsouyanni et al.. 2009) also analyzed the Canadian and
European datasets (the U.S. data were analyzed for PMi0 only) for evidence of a
threshold, using the threshold analysis method (Model 3) applied in Bell et al. (2006)
study described above. There was no evidence of a threshold in the Canadian data
(i.e., the pattern of AIC values for each increment of a potential threshold value
varied across cities, most of which showed no local minima). Likewise, the threshold
analysis conducted using the European data also showed no evidence of a threshold.
The PAPA study, did not examine whether a threshold exists in the O3-mortality C-R
relationship, but instead the shape of the C-R curve individually for each city
(Bangkok, Hong Kong, Shanghai, and Wuhan) (Wong et al.. 2010). Using a natural
spline smoother with 3df for the O3 term, Wong et al. (2010) examined whether non-
linearity was present by testing the change in deviance between the smoothed, non-
linear model and an unsmoothed linear model with 1 df For each of the cities, both
across the full range of the O3 distribution and specifically within the range of the
25th to 75th percentile of each city's 24-h avg O3 concentrations (i.e., a range of
9.7 ppb to 60.4 ppb across the cities) there was no evidence of a non-linear
relationship in the O3-mortality C-R curve. It should be noted that the range of the
25th to 75th percentiles of O3 concentrations in all of the cities, except Wuhan, was
at the lower end of the distribution observed in the U.S. using all-year data, where the
range from the 25th to 75th percentiles is 30 ppb to 50 ppb (Table 3-6).
Additional threshold analyses were conducted using NMMAPS data, by Xia and
Tong (2006) and Stylianou and Nicolich (2009). Both studies used a new statistical
approach developed by Xia and Tong (2006) to examine thresholds in the O3
mortality C-R relationship. The approach consisted of an extended GAM model,
which accounted for the cumulative and nonlinear effects of air pollution using a
weighted cumulative sum for each pollutant, with the weights (non-increasing further
into the past) derived by a restricted minimization method. The authors did not use
the term distributed lag model, but their model has the form of distributed lag model,
except that it allows for nonlinear functional forms. Using NMMAPS data for 1987-
1994 for 3 U.S. cities (Chicago, Pittsburgh, and El Paso), Xia and Tong (2006) found
that the extent of cumulative effects of O3 on mortality were relatively short. While
the authors also note that there was evidence of a threshold effect around 24-h avg
concentrations of 25 ppb, the threshold values estimated in the analysis were
sometimes in the range where data density was low. Thus, this threshold analysis
needs to be replicated in a larger number of cities to confirm this observation.
It should be noted that the model used in this analysis did not include a smooth
function of days to adjust for unmeasured temporal confounders, and instead adjusted
for season using a temperature term. As a result, these results need to be viewed with
caution because some potential temporal confounders (e.g., influenza) do not always
follow seasonal patterns of temperature.
6-256
-------
Stylianou and Nicolich (2009) examined the existence of thresholds following an
approach similar to Xia and Tong (2006) for all-cause, cardiovascular, and
respiratory mortality using data from NMMAPS for nine major U.S. cities
(i.e., Baltimore, MD, Chicago, IL, Dallas/Fort Worth,TX, Los Angeles, CA, Miami,
FL, New York, NY, Philadelphia, PA, Pittsburgh, PA, and Seattle, WA) for the years
1987-2000. The authors found that PMi0 and O3 were the two important predictors
of mortality. Stylianou and Nicolich (2009) found that the estimated O3-mortality
risks varied across the nine cities with the models exhibiting apparent thresholds, in
the 10-45 ppb range for O3 (3-day accumulation). However, given the city-to-city
variation in risk estimates, combining the city-specific estimates into an overall
estimate complicates the interpretation of a threshold. Unlike the Xia and Tong
(2006) analysis, Stylianou and Nicolich (2009) included a smooth function of time to
adjust for seasonal/temporal confounding, which could explain the difference in
results between the two studies.
In conclusion, the evaluation of the O3-mortality C-R relationship did not find any
evidence that supports a threshold in the relationship between short-term exposure to
O3 and mortality within the range of O3 concentrations observed in the United States.
Additionally, recent evidence suggests that the shape of the O3-mortality C-R curve
remains linear across the full range of O3 concentrations. However, the studies
evaluated demonstrated that the heterogeneity in the O3-mortality relationship across
cities (or regions) complicates the interpretation of a combined C-R curve and
threshold analysis. Given the effect modifiers identified in the mortality analyses that
are also expected to vary regionally (e.g., temperature, air conditioning prevalence), a
national or combined analysis may not be appropriate to identify whether a threshold
exists in the O3-mortality C-R relationship. Overall, the studies evaluated support a
linear O3-mortality C-R relationship and continue to support the conclusions from
the 2006 O3 AQCD, which stated that "if a population threshold level exists in O3
health effects, it is likely near the lower limit of ambient O3 concentrations in the
United States" (U.S. EPA. 2006b).
6.6.2.4 Associations of Cause-Specific Mortality and Short-term
O3 Exposure
In the 2006 O3 AQCD, an evaluation of studies that examined cause-specific
mortality found consistent positive associations between short-term O3 exposure and
cardiovascular mortality, with less consistent evidence for associations with
respiratory mortality. The majority of the evidence for associations between O3
exposure and cause-specific mortality were from single-city studies, which had small
daily mortality counts and subsequently limited statistical power to detect
associations.
New multicity studies evaluated in this review build upon and confirm the
associations between short-term O3 exposure and cause-specific mortality identified
in the 2006 O3 AQCD (U.S. EPA. 2006b) (Figure 6-37 [and Table 6-531).
6-257
-------
In APHENA, a multicontinent study that consisted of the NMMAPS, APHEA2 and
Canadian multicity datasets, consistent positive associations were reported for both
cardiovascular and respiratory mortality in all-year analyses when focusing on the
natural spline model with 8 df/year (Figure 6-37 [and Table 6-531). The associations
between O3 exposure and cardiovascular and respiratory mortality in all-year
analyses were further supported by the multicity PAPA study (Wong et al.. 2010).
The magnitude of cardiovascular mortality associations were primarily larger in
analyses restricted to the summer season compared to those observed in all-year
analyses (Figure 6-37 [and Table 6-531). Additional multicity studies from the U.S.
(Zanobetti and Schwartz. 2008b) and Europe (Stafoggia et al.. 2010: Samoli et al..
2009) that conducted summer season analyses provide evidence supporting
associations between O3 exposure and cardiovascular and respiratory mortality that
are similar or larger in magnitude compared to those observed in all-year analyses.
Of the studies evaluated, only the APHENA study (Katsouyanni et al., 2009) and an
Italian multicity study (Stafoggia et al., 2010) conducted an analysis of the potential
for copollutant confounding of the O3 cause-specific mortality relationship. When
focusing on the natural spline model with 8 df/year and lag 1 results (as discussed in
Section 6.6.2.1), the APHENA study found that O3 cause-specific mortality risk
estimates were fairly robust to the inclusion of PMi0 in copollutant models in the
European dataset with more variability in the U.S. and Canadian datasets
(i.e., copollutant risk estimates increased and decreased for respiratory and
cardiovascular mortality). In summer season analyses cardiovascular O3 mortality
risk estimates were robust in the European dataset and attenuated but remained
positive in the U.S. datasets; whereas, respiratory O3 mortality risk estimates were
attenuated in the European dataset and robust in the U.S. dataset. The authors did not
examine copollutant models during the summer season in the Canadian dataset
(Figure 6-30 [and Table 6-451). Interpretation of these results requires caution;
however, due to the different PM sampling schedules employed in each of these
study locations (i.e., primarily every-6th day in the U.S. and Canadian datasets and
every-day in the European dataset). The results of the summer season analyses from
the APHENA study (Katsouvanni et al.. 2009) are consistent with those from a study
of 10 Italian cities during the summer months (Stafoggia et al.. 2010). Stafoggia et al.
(2010) found that cardiovascular (14.3% [95% CI: 6.7, 22.4%]) and cerebrovascular
(8.5% [95% CI: 0.06, 16.3%]) mortality O3 effect estimates were robust to the
inclusion of PM10 in copollutant models (14.3% [95% CI: 6.7, 23.1%] and 7.3%
[95% CI: -1.2, 16.3], respectively), while respiratory mortality O3 effects estimates
(17.6% [95% CI: 1.8, 35.5%]) were attenuated, but remained positive (9.2%
[95% CI: -6.9,28.8%]).
6-258
-------
Study
Bell etal. (2005)3
Wong eta I. (2010)
Katsouyanni etal. (2009)
Gryparisetal. (2004)a
Samolietal. (2009)
Zanobetti and Schwartz (2008)
Stafoggiaetal. (2010)
Katsouyanni etal. (2009)
Bell etal. (2005)a
Wong eta I. (2010)
Katsouyanni etal. (2009)
Gryparisetal. (2004)a
Zanobetti and Schwartz (2008)
Katsouyanni etal. (2009)
Samolietal. (2009)
Stafoggiaetal. (2010)
Katsouyanni etal. (2009)
Ages
Lag
U.S. and non-U.S.
PAPA (4 cities]
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
21 European cities
21 European cities
48U.S. cities
lOltaliancities
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
U.S. and non-U.S.
PAPA (4 cities]
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
21 European cities
48U.S. cities
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
21 European cities
lOltaliancities
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
All
275
<75
All
235
275
<75
All
275
All
235
275
NR
0-1
DL(0-2
DL(0-2
DL(0-2]
DL(0-2
DL(0-2
DL(0-2
DL(0-2]
DL(0-2
0-1
0-1
0-3
DL(0-5
DL(0-2
DL(0-2
DL(0-2]
DL(0-2
DL(0-2
DLJO-2
DL(0-2]
DL(0-2
NR
0-1
DL(0-2
DL(0-2
DL(0-2]
DL(0-2
DL(0-2
DLJO-2
DL(0-2]
DL(0-2
0-1
0-3
DL(0-2
DL(0-2
DL(0-2]
DL(0-2
0-1
DL(0-5
DLJO-2
DL(0-2
DL(0-2]
DL(0-2
Cardiovascular
-o-
Respiratory
-10 -5
Effect estimates are for a 20 ppb increase in 24-h avg; 30 in 8-h max; and 40ppb increase in 1-h max O3 concentrations.
Red = cardiovascular; blue = respiratory; closed circles = all-year analysis; and open circles = summer-only analysis. An "a"
represents studies from the 2006 O3 AQCD. A "b" represents risk estimates from APHENA-Canada standardized to an
approximate IQR of 5.1 ppb for a 1-h max increase in O3 concentrations (Section 6.2.7.2).
Figure 6-37 Percent increase in cause-specific mortality.
6-259
-------
Table 6-53 Corresponding effect estimates for Figure
Study*
Location
Ages Lag
6-37.
Avg Time %lncrease (95% Cl)
Cardiovascular
All-year - Cardiovascular
Bell et al. (2005)a
Wongetal. (2010)
Katsouvannietal. (2009)
U.S. and non-U.S.
PAPA (4 cities)
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
All NR
0-1
> 75 DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
<75 DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
24-h avg 2.23 (1 .36,3.08)
8-h max 2.20 (0.06, 4.37)
1-hmax 2.30 (-1 .33, 6.04)
8.96(0.75,18.6)
1.1 (0.10,2.20)
2.06 (-0.24, 4.31)
3.83 (-0.1 6, 7.95)
7.03 (-2.71, 17.7)
0.87 (-0.35, 2.10)
1.98 (-1.09, 5.13)
Summer - Cardiovascular
Gryparisetal. (2004)a
Samoli et al. (2009)
Zanobetti and Schwartz (2008b)
Stafoggia et al. (2010)
Katsouvannietal. (2009)
21 European cities
21 European cities
48 U.S. cities
10 Italian cities
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
All 0-1
0-1
0-3
> 35 DL(0-5)
> 75 DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
<75 DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
8-h max 2.7(1.29,4.32)
8-h max 1.48(0.18,2.80)
8-h max 2.42 (1 .45, 3.43)
8-h max 14.3(6.65,22.4)
1-hmax 3. 18 (-0.47, 6.95)
1 .50 (-2.79, 5.95)
0.1 9 (-0.36, 0.74)
3.67 (0.95, 6.53)
6.78(2.70, 11.0)
-1.02 (-4.23, 2.30)
-0.1 3 (-0.55, 0.29)
2.22 (-1 .48, 6.04)
Respiratory
All-years - Respiratory
Bell et al. (2005)a
Wongetal. (2010)
Katsouvannietal. (2009)
U.S. and non-U.S.
PAPA (4 cities)
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Eeurope
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
All NR
0-1
DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
> 75 DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
24-h avg 0.94 (-1 .02, 2.96)
8-h max 2.02 (-0.41 , 4.49)
1-hmax 2.54 (-3.32, 8.79)
1.02 (-11. 9, 15.9)
0.1 3 (-1.60, 1.90)
1.82 (-2. 18, 6.04)
1.10 (-6.48, 9.21)
-4.61 (-19.3, 13.3)
-0.60 (-2.70, 1 .60)
1.10 (-3.48, 5.95)
6-260
-------
Study*
Location
Ages Lag
Avg Time
"/.Increase (95% Cl)
Summer - Respiratory
Grypariset al. (2004)a
Zanobetti and Schwartz (2008b)
Katsouvanni et al. (2009)
Samoli et al. (2009)
Stafoggia et al. (2010)
Katsouvanni et al. (2009)
21 European cities
48 U.S. cities
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
21 European cities
10 Italian cities
APHENA-U.S.
APHENA-Canada
APHENA-Canada
APHENA-Europe
All 0-1
0-3
DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
0-1
> 35 DL(0-5)
> 75 DL(0-2)
DL(0-2)
DL(0-2)b
DL(0-2)
8-h max
8-h max
1-h max
8-h max
8-h max
1-h max
6.75(4.38,9.10)
2.51 (1.14, 3.89)
4.40 (-2. 10, 11.3)
26.1 (13.3, 41.2)
3.00 (1 .60, 4.50)
3.83 (-1.33, 9.21)
2.38(0.65,4.19)
17.6(1.78,35.5)
4.07 (-4.23, 13.0)
19.5(2.22, 40.2)
2.30 (0.28, 4.40)
2.46 (-3.40, 8.62)
'Studies from Figure 6-37. plus others.
"Studies from the 2006 O3 AQCD.
bRisk estimates from APHENA-Canada standardized to an approximate IQR of 5.1 ppb for a 1-h max increase in O3 concentrations
(Section 6.2.7.2).
Collectively, the results from the new multicity studies provide evidence of
associations between short-term O3 exposure and cardiovascular and respiratory
mortality with additional evidence indicating these associations persist, and in some
cases the magnitude of associations are increased, in the summer season. Although
copollutant analyses of cause-specific mortality are limited, the APHENA study
found that O3 cause-specific mortality risk estimates were fairly robust to the
inclusion of PMi0 in copollutant models when focusing on the dataset with daily
PMio data (i.e., the European dataset), which is supported by the results from
Stafoggia et al. (2010). Additionally, APHENA found that O3 cause-specific
mortality risk estimates were moderately to substantially sensitive (e.g., increased or
attenuated) to inclusion of PMi0 in the U.S. and Canadian datasets. However, the
mostly every-6th-day sampling schedule for PMi0 in the U.S. and Canadian datasets
greatly reduced their sample size and limits the interpretation of these results.
6.6.3 Summary and Causal Determination
The evaluation of new multicity studies that examined the association between short-
term O3 exposure and mortality found evidence which supports the conclusions of
the 2006 O3 AQCD. These new studies reported consistent positive associations
between short-term O3 exposure and all-cause (nonaccidental) mortality, with
associations persisting or increasing in magnitude during the warm season, and
provide additional support for associations between O3 exposure and cardiovascular
and respiratory mortality.
6-261
-------
Recent studies further examined potential confounders (e.g., copollutants and
seasonality) of the O3-mortality relationship. Because the PM-O3 correlation varies
across regions, due to the difference in PM chemical constituents, interpretation of
the combined effect of PM on the relationship between O3 and mortality is not
straightforward. Unlike previous studies that were limited to primarily examining the
confounding effects of PMi0, the new studies expanded their analyses to include
multiple PM indices (e.g., PMi0, PM2.5, and PM components). An examination of
copollutant models found evidence that associations between O3 and all-cause
mortality were robust to the inclusion of PMi0 or PM2.5 (Stafoggia et al.. 2010:
Katsouvanni et al.. 2009: Bell et al.. 2007). while other studies found evidence for a
modest reduction (-20-30%) when examining PMi0 (Smith et al.. 2009b). Additional
evidence suggests potential sensitivity (e.g., increases and attenuation) of O3
mortality risk estimates to copollutants by age group or cause-specific mortality
(e.g., respiratory and cardiovascular) (Stafoggia et al.. 2010: Katsouvanni et al..
2009). An examination of PM components, specifically sulfate, found evidence for
reductions in O3-mortality risk estimates in copollutant models (Franklin and
Schwartz. 2008). Overall, across studies, the potential impact of PM indices on
O3-mortality risk estimates tended to be much smaller than the variation in
O3-mortality risk estimates across cities suggesting that O3 effects are independent of
the relationship between PM and mortality. However, interpretation of the potential
confounding effects of PM on O3-mortality risk estimates requires caution. This is
because the PM-O3 correlation varies across regions, due to the difference in PM
components, complicating the interpretation of the combined effect of PM on the
relationship between O3 and mortality. Additionally, the limited PM or PM
component datasets used as a result of the every-3rd- and 6th-day PM sampling
schedule instituted in most cities limits the overall sample size employed to examine
whether PM or one of its components confounds the O3-mortality relationship.
An examination of potential seasonal confounding of the O3-mortality relationship
found that the extent of smoothing or the methods used for adjustment can influence
O3 risk estimates when not applying enough degrees of freedom to control for
temporal/season trends (Katsouvanni et al.. 2009). This is because of the opposing
seasonal trends between O3 and mortality.
The multicity studies evaluated within this section also examined the regional
heterogeneity observed in O3-mortality risk estimates. These studies provide
evidence which suggests generally higher O3-mortality risk estimates in northeastern
U.S. cities with some regions showing no associations between O3 exposure and
mortality (e.g., Southwest, Urban Midwest) (Smith et al.. 2009b: Bell and Dominici,
2008). Multicity studies that examined individual- and community-level
characteristics identified characteristics that may explain the observed regional
heterogeneity in O3-mortality risk estimates as well as characteristics of populations
potentially at greatest risk for O3-related health effects. An examination of
community-level characteristics found an increase in the O3-mortality risk estimates
in cities with higher unemployment, percentage of the population Black/African-
American, percentage of the working population that uses public transportation,
lower temperatures, and lower prevalence of central air conditioning (Medina-Ramon
6-262
-------
and Schwartz, 2008). Additionally, a potential interactive, or synergistic, effect on
the O3-mortality relationship was observed when examining differences in the
O3-mortality association across temperature levels (Ren et al., 2008).
An examination of individual-level characteristics found evidence that older age,
female sex, Black race, having atrial fibrillation, SES indicators (i.e., educational
attainment, income level, and employment status), and out-of hospital deaths,
specifically in those individuals with diabetes, modify O3-mortality associations
(Cakmaketal.. 2011: Stafoggiaet al.. 2010: Medina-Ramon and Schwartz. 2008).
and lead to increased risk of O3-related mortality. Overall, additional research is
warranted to further confirm whether these characteristics, individually or in
combination, can explain the observed regional heterogeneity.
Additional studies were evaluated that examined factors that may influence the shape
of the O3-mortality C-R curve, such as multi-day effects, mortality displacement,
adaptation, the use of different exposure metrics (i.e., 24-h avg, 8-h max or 1-h max),
and whether a threshold exists in the O3-mortality relationship. An examination of
multiday effects in a U.S. and European multicity study found conflicting evidence
for mortality displacement, but both studies suggest that the positive associations
between O3 and mortality are observed mainly in the first few days after exposure
(Samoli et al.. 2009: Zanobetti and Schwartz. 2008b). A U.S. multicity study found
evidence of an adaptive response to O3 exposure, with the highest risk estimates
earlier in the O3 season (i.e., July) and diminished effects later (i.e., August)
(Zanobetti and Schwartz. 2008a). However, the evidence of adaptive effects has an
implication for the interpretation of multi-day effects, and requires further analysis.
The limited number of studies conducted that examined the effect of using different
exposure metrics (i.e., 1-h max, 8-h max, and 24-h avg) when examining the O3-
mortality relationship found relatively comparable O3-mortality risk estimates across
the exposure metrics used (Smith et al.. 2009b: Gryparis et al.. 2004). Analyses that
specifically focused on the O3-mortality C-R relationship supported a linear O3-
mortality relationship and found no evidence of a threshold within the range of O3
concentrations in the U.S., but did observe evidence for potential differences in the
C-R relationship across cities (Katsouvanni et al.. 2009: Stylianou and Nicolich.
2009: Bell et al.. 2006). Collectively, these studies support the conclusions of the
2006 O3 AQCD that "if a population threshold level exists in O3 health effects, it is
likely near the lower limit of ambient O3 concentrations in the U.S."
Studies that examined the association between short-term O3 exposure and cause-
specific mortality confirm the associations with both cardiovascular and respiratory
mortality reported in the 2006 O3 AQCD (Stafoggia et al.. 2010: Wong et al.. 2010:
Katsouvanni et al.. 2009: Samoli et al.. 2009: Zanobetti and Schwartz. 2008b). These
associations were primarily larger in magnitude during the summer season compared
to all-year analyses. Of the studies that examined the potential confounding effects of
PM [i.e., Stafoggia et al. (2010): Katsouvanni et al. (2009)1. O3 mortality
associations remained relatively robust in copollutant models, but interpretation of
these studies was complicated by the different PM sampling schedules (e.g., every -
6th-day) employed in each study. Overall, the strong evidence for respiratory effects
due to short-term O3 exposure (Section 6.2) are consistent across disciplines and
6-263
-------
provides coherence for the respiratory mortality associations observed across studies.
The strong evidence for O3-induced cardiovascular mortality is supported by
controlled human exposure and animal toxicological studies that provide initial
evidence for a biologically plausible mechanism for O3-induced cardiovascular
mortality. However, a lack of coherence with epidemiologic studies of cardiovascular
morbidity that do not demonstrate consistent evidence of O3-induced cardiovascular
effects complicate the evidence for a biological pathway of events leading to
mortality (Section 6.3).
In conclusion, the recent epidemiologic studies build upon and confirm the
associations between short-term O3 exposure and all-cause and cause-specific
mortality reported in the 2006 O3 AQCD. However, there is a lack of coherence
across disciplines and consistency across health outcomes for O3-induced
cardiovascular morbidity (Section 6.3) which do not support the relatively strong
epidemiologic evidence for O3-related cardiovascular mortality. Overall, recent
studies have provided additional information regarding key uncertainties (previously
identified - including the potential confounding effects of copollutants and seasonal
trend), individual- and community-level factors that may lead to increased risk of
O3-induced mortality and the heterogeneity in O3-mortality risk estimates, and
continued evidence of a linear no-threshold C-R relationship. Although some
uncertainties still remain, the collective body of evidence is sufficient to conclude
there is likely to be a causal relationship between short-term O3 exposure and
total mortality.
6.7 Overall Summary
The evidence reviewed in this chapter describes the recent findings regarding the
health effects of short-term exposure to ambient O3 concentrations. Table 6-54
provides an overview of the causal determinations for each of the health categories
evaluated.
Table 6-54 Summary of causal determinations for short-term exposures to O3.
Health Category
Causal Determination
Respiratory Effects
Causal relationship
Cardiovascular Effects
Likely to be a causal relationship
Central Nervous System Effects
Suggestive of a causal relationship
Effects on Liver and Xenobiotic Metabolism
Inadequate to infer a causal relationship
Effects on Cutaneous and Ocular Tissues
Inadequate to infer a causal relationship
Total Mortality
Likely to be a causal relationship
6-264
-------
References
Adamkiewicz. G: Ebelt. S: Syring. M; Slater. J; Speizer. FE; Schwartz. J; Suh. H; Gold. DR. (2004).
Association between air pollution exposure and exhaled nitric oxide in an elderly population. Thorax 59:
204-209. http://dx.doi.org/10.n36/thorax.2003.006445
Adams. WC. (1998). Dose-response effect of varied equivalent minute ventilation rates on pulmonary function
responses during exposure to ozone. Washington, DC: American Petroleum Institute.
Adams. WC. (2000). Ozone dose-response effects of varied equivalent minute ventilation rates. J Expo Sci
Environ Epidemiol 10: 217-226. http://dx.doi.org/10.1038/sj.jea.7500086
Adams. WC. (2002). Comparison of chamber and face-mask 6.6-hour exposures to ozone on pulmonary
function and symptoms responses. Inhal Toxicol 14: 745-764.
http://dx.doi.org/10.1080/08958370290084610
Adams. WC. (2003a). Comparison of chamber and face mask 6.6-hour exposure to 0.08 ppm ozone via square-
wave and triangular profiles on pulmonary responses. Inhal Toxicol 15: 265-281.
Adams. WC. (2003b). Relation of pulmonary responses induced by 66-h exposures to 0.08 ppm ozone and 2-h
exposures to 0.30 ppm ozone via chamber and face-mask inhalation. Inhal Toxicol 15: 745-759.
Adams. WC. (2006a). Comparison of chamber 6.6-h exposures to 0.04-0.08 ppm ozone via square-wave and
triangular profiles on pulmonary responses. Inhal Toxicol 18: 127-136.
http://dx.doi.org/10.1080/08958370500306107
Adams. WC. (2006b). Human pulmonary responses with 30-minute time intervals of exercise and rest when
exposed for 8 hours to 0.12 ppm ozone via square-wave and acute triangular profiles. Inhal Toxicol 18:
413-422. http://dx.doi.org/10.1080/08958370600563599
Adams. WC: Ollison. WM. (1997). Effects of prolonged simulated ambient ozone dosing patterns on human
pulmonary function and symptomatology. Pittsburgh, PA: Air & Waste Management Association.
Adams. WC: Schelegle. ES. (1983). Ozone and high ventilation effects on pulmonary function and endurance
performance. JAppl Physiol 55: 805-812.
Aibo. PI; Birmingham. NP; Lewandowski. R; Maddox. JF; Roth. RA; Ganev. PE; Wagner. JG: Harkema. JR.
(2010). Acute exposure to ozone exacerbates acetaminophen-induced liver injury in mice. Toxicol Sci 115:
267-285. http://dx.doi.org/10.1093/toxsci/kfq034
Alexeeff. SE; Litonjua. AA; Suh. H; Sparrow. D; Vokonas. PS: Schwartz. J. (2007). Ozone exposure and lung
function: Effect modified by obesity and airways hyperresponsiveness in the VA Normative Aging Study.
Chest 132: 1890-1897. http://dx.doi.org/10.1378/chest.07-1126
Alexeeff. SE; Litonjua. AA; Wright. RO: Baccarelli. A; Suh. H; Sparrow. D; Vokonas. PS: Schwartz. J. (2008).
Ozone exposure, antioxidant genes, and lung function in an elderly cohort: VA Normative Aging Study.
Occup Environ Med 65: 736-742. http://dx.doi.org/10.1136/oem.2007.035253
Alexis. N; Urch. B; Tarlo. S: Corey. P; Pengelly. D: O'Byrne. P; Silverman. F. (2000). Cyclooxygenase
metabolites play a different role in ozone-induced pulmonary function decline in asthmatics compared to
normals. Inhal Toxicol 12: 1205-1224.
Alexis. NE; Lav. JC: Hazucha. M; Harris. B; Hernandez. ML; Bromberg. PA; Kehrl. H; Diaz-Sanchez. D;
Kim. C: Devlin. RB; Peden. DB. (2010). Low-level ozone exposure induces airways inflammation and
modifies cell surface phenotypes in healthy humans. Inhal Toxicol 22: 593-600.
http://dx.doi.org/10.3109/08958371003596587
6-265
-------
Alexis. NE; Zhou. H; Lav. JC: Harris. B; Hernandez. ML; Lu. TS: Bromberg. PA; Diaz-Sanchez. D; Devlin.
RB; Kleeberger. SR; Peden. DB. (2009). The glutathione-S-transferase Mu 1 null genotype modulates
ozone-induced airway inflammation in human subjects. J Allergy Clin Immunol 124: 1222-1228.
http://dx.doi.0rg/10.1016/i.jaci.2009.07.036
Alfaro-Rodriguez. A; Gonzalez-Pina. R. (2005). Ozone-induced paradoxical sleep decrease is related to
diminished acetylcholine levels in the medial preoptic area in rats. Chem Biol Interact 151: 151-158.
http://dx.doi.0rg/10.1016/i.cbi.2004.10.001
Anderson. HR; Armstrong. B; Hajat. S: Harrison. R; Monk. V; Poloniecki. J; Timmis. A; Wilkinson. P. (2010).
Air pollution and activation of implantable cardioverter defibrillators in London. Epidemiology 21: 405-
413. http://dx.doi.org/10.1097/EDE.Ob013e3181d61600
Angoa-Perez. M; Jiang. H; Rodriguez. AI; Lemini. C: Levine. RA; Rivas-Arancibia. S. (2006). Estrogen
counteracts ozone-induced oxidative stress and nigral neuronal death. Neuroreport 17: 629-633.
Apte. MG: Buchanan. IS: Mendell. MJ. (2008). Outdoor ozone and building-related symptoms in the BASE
study. Indoor Air 18: 156-170. http://dx.doi.Org/10.llll/i.1600-0668.2008.00521.x
Araneda. S: Commin. L; Atlagich. M; Kitahama. K; Parraguez. VH; Pequignot. JM; Dalmaz. Y. (2008). VEGF
overexpression in the astroglial cells of rat brainstem following ozone exposure. Neurotoxicology 29: 920-
927. http://dx.doi.0rg/10.1016/j.neuro.2008.09.006
Aranyi. C: Vana. SC: Thomas. PT; Bradof. JN; Fenters. JD; Graham. JA; Miller. FJ. (1983). Effects of
subchronic exposure to a mixture of O3, SO2, and (NH4)2SO4 on host defenses of mice. J Toxicol Environ
Health 12: 55-71. http://dx.doi.org/10.1080/15287398309530407
Arbex. AM; de Souza Conceicao. GM; Perez Cendon. S: Arbex. FF; Lopes. AC: Providello Moyses. E;
Santiago. SL; Saldiva. PHN; Pereira. LAA; Ferreira Braga. AL. (2009). Urban air pollution and COPD-
related emergency room visits. J Epidemiol Community Health 966: 777-783.
http://dx.doi.org/10.1136/iech.2008.078360
Aris. RM; Tager. I; Christian. D; Kelly. T; Balmes. JR. (1995). Methacholine responsiveness is not associated
with 03-induced decreases in FEV1. Chest 107: 621-628.
Arito. H; Takahashi. M; Iwasaki. T; Uchiyama. I. (1997). Age-related changes in ventilatory and heart rate
responses to acute ozone exposure in the conscious rat. Ind Health 35: 78-86.
Arito. H; Uchiyama. I; Arakawa. H; Yokoyama. E. (1990). Ozone-induced bradycardia and arrhythmia and
their relation to sleep-wakefulness in rats. Toxicol Lett 52: 169-178. http://dx.doi.org/10.1016/0378-
4274(90)90151-6
Arito. H; Uchiyama. I; Yokoyama. E. (1992). Acute effects of ozone on EEG activity, sleep-wakefulness and
heart rate in rats. Ind Health 30: 23-34.
Armstrong. BG. (2003). Fixed factors that modify the effects of time-varying factors: Applying the case-only
approach. Epidemiology 14: 467-472. http://dx.doi.org/10.1097/01.ede.0000071408.39011.99
Atkinson. RW: Bremner. SA; Anderson. HR; Strachan. DP; Bland. JM; Ponce de Leon. A. (1999). Short-term
associations between emergency hospital admissions for respiratory and cardiovascular disease and outdoor
air pollution in London. Arch Environ Occup Health 54: 398-411.
ATS (American Thoracic Society). (1991). Lung function testing: Selection of reference values and
interpretative strategies. Am J Respir Crit Care Med 144: 1202-1218.
ATS (American Thoracic Society). (2000a). Guidelines for methacholine and exercise challenge testing-1999.
Am J Respir Crit Care Med 161: 309-329.
Avila-Costa. MR; Colin-Barenque. L; Fortoul. TI; Machado-Salas. JP; Espinosa-Villanueva. J: Rugerio-
Vargas. C: Rivas-Arancibia. S. (1999). Memory deterioration in an oxidative stress model and its
correlation with cytological changes on rat hippocampus CA1. Neurosci Lett 270: 107-109.
6-266
-------
Avissar. NE; Reed. CK; Cox. C: Frampton. MW: Finkelstein. JN. (2000). Ozone, but not nitrogen dioxide,
exposure decreases glutathione peroxidases in epithelial lining fluid of human lung. Am J Respir Crit Care
Med 162: 1342-1347.
Avol. EL; Linn. WS: Venet. TG: Shamoo. DA; Hackney. JD. (1984). Comparative respiratory effects of ozone
and ambient oxidant pollution exposure during heavy exercise. J Air Waste Manag Assoc 34: 804-809.
Avol. EL; Navidi. WC: Rappaport. EB; Peters. JM. (1998b). Acute effects of ambient ozone on asthmatic,
wheezy, and healthy children. (82). Topsfield, MA: Health Effects Institute; Flagship Press.
Avol. EL; Trim. SC: Little. DE; Spier. CE; Smith. MN; Peng. RC: Linn. WS: Hackney. JD: Gross. KB; D'Arcv.
JB; Gibbons. D; Higgins. ITT. (1990). Ozone exposure and lung function in children attending a southern
California summer camp. In Proceedings of the 83rd A&WMA Annual Meeting. Pittsburgh, PA: Air &
Waste Management Association.
Azevedo. JM; Goncalves. FL; de FatimaAndrade. M. (2011). Long-range ozone transport and its impact on
respiratory and cardiovascular health in the north of Portugal. Int J Biometeorol 55: 187-202.
http://dx.doi.org/10.1007/s00484-010-0324-2
Baccarelli. A; Zanobetti. A; Martinelli. I; Grillo. P; Hou. L; Lanzani. G: Mannucci. PM; Bertazzi. PA;
Schwartz. J. (2007). Air pollution, smoking, and plasma homocysteine. Environ Health Perspect 115: 176-
181. http://dx.doi.org/10.1289/ehp.9517
Baccini. M; Biggeri. A; Accetta. G: Kosatsky. T; Katsouvanni. K; Analitis. A; Anderson. HR; Bisanti. L;
D'Ippoliti. D; Danova. J: Forsberg. B; Medina. S: Paldv. A; Rabczenko. D; Schindler. C: Michelozzi. P.
(2008). Heat effects on mortality in 15 European cities. Epidemiology 19: 711-719.
http://dx.doi.org/10.1097/EDE.Ob013e318176bfcd
Baja. ES: Schwartz. JD: Wellenius. GA; Coull. BA; Zanobetti. A; Vokonas. PS: Suh. HH. (2010). Traffic-
related air pollution and QT interval: Modification by diabetes, obesity, and oxidative stress gene
polymorphisms in the Normative Aging Study. Environ Health Perspect 118: 840-846.
http://dx.doi.org/10.1289/ehp.0901396
Balbi. B; Pignatti. P; Corradi. M; Baiardi. P; Bianchi. L; Brunetti. G: Radaeli. A; Moscato. G: Mutti. A;
Spanevello. A; Malerba. M. (2007). Bronchoalveolar lavage, sputum and exhaled clinically relevant
inflammatory markers: Values in healthy adults [Review]. Eur Respir J 30: 769-781.
http://dx.doi.org/10.1183/09031936.00112306
Ballester. F; Rodriguez. P; Iniguez. C: Saez. M; Daponte. A; Galan. I; Taracido. M; Arribas. F; Bellido. J:
Cirarda. FB; Canada. A; Guillen. JJ; Guillen-Grima. F; Lopez. E; Perez-Hoyos. S: Lertxundi. A; Toro. S.
(2006). Air pollution and cardiovascular admisisons association in Spain: Results within the EMECAS
project. J Epidemiol Community Health 60: 328-336.
Ballester. F; Saez. M; Daponte. A; Ordonez. JM: Taracido. M; Cambra. K; Arribas. F; Bellido. JB; Guillen. JJ:
Aguinaga. I; Canada. A; Lopez. E; Iniguez. C: Rodriguez. P; Perez-Hoyos. S: Barcelo. MA; Ocana. R;
Aranguez. E. (2005). [The EMECAS Project: Spanish multicentre study on short-term health effects of air
pollution]. Rev Esp Salud Publica 79: 229-242.
Balmes. JR; Arts. RM; Chen. LL; Scannell. C: Tager. IB; Finkbeiner. W: Christian. D; Kelly. T; Hearne. PQ:
Ferrando. R; Welch. B. (1997). Effects of ozone on normal and potentially sensitive human subjects. Part I:
Airway inflammation and responsiveness to ozone in normal and asthmatic subjects (pp. 1-37; discussion
81-99). (ISSN 1041-5505). Boston, MA: Health Effects Institute.
Balmes. JR; Chen. LL; Scannell. C; Tager. I; Christian. D; Hearne. PQ; Kelly. T; Aris. RM. (1996). Ozone-
induced decrements in FEV1 and FVC do not correlate with measures of inflammation. Am J Respir Crit
Care Med 153: 904-909.
Barnes. PJ; Liew. FY. (1995). Nitric oxide and asthmatic inflammation [Review]. Immunol Today 16: 128-130.
http://dx.doi.org/10.1016/0167-5699(95)80128-6
Barnett. AG; Williams. GM; Schwartz. J; Best. TL; Neller. AH; Petroeschevsky. AL; Simpson. RW (2006).
The effects of air pollution on hospitalizations for cardiovascular disease in elderly people in Australian and
New Zealand cities. Environ Health Perspect 114: 1018-1023. http://dx.doi.org/10.1289/ehp.8674
6-267
-------
Barraza-Villarreal. A; Sunyer. J; Hernandez-Cadena. L; Escamilla-Nunez. MC: Sienra-Monge. JJ: Ramirez-
Aguilar. M; Cortez-Lugo. M; Holguin. F; Diaz-Sanchez. D; Olin. AC: Romieu. I. (2008). Air pollution,
airway inflammation, and lung function in a cohort study of Mexico City schoolchildren. Environ Health
Perspect 116: 832-838. http://dx.doi.org/10.1289/ehp.10926
Basha. MA; Gross. KB; Gwizdala. CJ: Haidar. AH; Popovich. J. Jr. (1994). Bronchoalveolar lavage
neutrophilia in asthmatic and healthy volunteers after controlled exposure to ozone and filtered purified air.
Chest 106: 1757-1765.
Bauer. AK; Rondini. EA; Hummel. KA; Degraff. LM; Walker. C: Jedlicka. AE; Kleeberger. SR. (2011).
Identification of candidate genes downstream of TLR4 signaling after ozone exposure in mice: A role for
heat shock protein 70. Environ Health Perspect 119: 1091-1097. http://dx.doi.org/10.1289/ehp.1003326
Bell. ML; Dominici. F. (2008). Effect modification by community characteristics on the short-term effects of
ozone exposure and mortality in 98 US communities. Am J Epidemiol 167: 986-997.
http://dx.doi.org/10.1093/aie/kwm396
Bell. ML; Dominici. F; Samet. JM. (2005). A meta-analysis of time-series studies of ozone and mortality with
comparison to the national morbidity, mortality, and air pollution study. Epidemiology 16: 436-445.
htrp://dx.doi.org/10.1097/01.ede.0000165817.40152.85
Bell. ML; Kim. JY; Dominici. F. (2007). Potential confounding of particulate matter on the short-term
association between ozone and mortality in multisite time-series studies. Environ Health Perspect 115:
1591-1595. http://dx.doi.org/10.1289/ehp.10108
Bell. ML; Lew. JK; Lin. Z. (2008). The effect of sandstorms and air pollution on cause-specific hospital
admissions in Taipei, Taiwan. Occup Environ Med 65: 104-111.
http://dx.doi.org/10.1136/oem.2006.031500
Bell. ML; McDermott. A; Zeger. SL; Samet. JM; Dominici. F. (2004). Ozone and short-term mortality in 95
US urban communities, 1987-2000. JAMA 292: 2372-2378. http://dx.doi.org/10.1001/iama.292.19.2372
Bell. ML; Peng. RD; Dominici. F. (2006). The exposure-response curve for ozone and risk of mortality and the
adequacy of current ozone regulations. Environ Health Perspect 114: 532-536.
Bennett. WD; Hazucha. MJ; Folinsbee. LJ; Bromberg. PA; Kissling. GE; London. SJ. (2007). Acute
pulmonary function response to ozone in young adults as a function of body mass index. Inhal Toxicol 19:
1147-1154. htrp://dx.doi.org/10.1080/08958370701665475
Bergamaschi. E; De Palma. G: Mozzoni. P; Vanni. S: Vettori. MV; Broeckaert. F; Bernard. A; Mutti. A.
(2001). Polymorphism of quinone-metabolizing enzymes and susceptibility to ozone-induced acute effects.
Am J Respir Crit Care Med 163: 1426-1431.
Berhane. K; Zhang. Y; Linn. WS: Rappaport. EB; Bastain. TM; Salam. MT; Islam. T; Lurmann. F; Gilliland.
FD. (2011). The effect of ambient air pollution on exhaled nitric oxide in the Children's Health Study. Eur
Respir J 37: 1029-1036. http://dx.doi.org/10.1183/09031936.00081410
Berkev. CS: Hoaglin. DC: Antczak-Bouckoms. A; Mosteller. F; Colditz. GA. (1998). Meta-analysis of multiple
outcomes by regression with random effects. Stat Med 17: 2537-2550.
http://dx.doi.org/10.1002/(SICI)1097-0258(19981130)17:22<2537::AID-SIM953>3.0.CO:2-C
Berry. M; Lioy. PJ; Gelperin. K; Buckler. G: Klotz. J. (1991). Accumulated exposure to ozone and
measurement of health effects in children and counselors at two summer camps. Environ Res 54: 135-150.
Biggeri. A; Baccini. M; Bellini. P; Terracini. B. (2005). Meta-analysis of the Italian studies of short-term
effects of air pollution (MISA), 1990-1999. Int J Occup Environ Health 11: 107-122.
Blomberg. A; Mudwav. IS: Nordenhall. C: Hedenstrom. H; Kelly. FJ; Frew. AJ; Holgate. ST; Sandstrom. T.
(1999). Ozone-induced lung function decrements do not correlate with early airway inflammatory or
antioxidant responses. Eur Respir J 13: 1418-1428.
Bosson. J; Stenfors. N; Bucht. A; Helledav. R; Pourazar. J: Holgate. ST; Kelly. FJ: Sandstrom. T; Wilson. S:
Frew. AJ; Blomberg. A. (2003). Ozone-induced bronchial epithelial cytokine expression differs between
healthy and asthmatic subjects. Clin Exp Allergy 33: 777-782.
6-268
-------
Boussouar. A; Araneda. S: Hamelin. C: Soulage. C: Kitahama. K; Dalmaz. Y. (2009). Prenatal ozone exposure
abolishes stress activation of Fos and tyrosine hydroxylase in the nucleus tractus solitarius of adult rat.
Neurosci Lett 452: 75-78.
Brauer. M; Blair. J; Vedal. S. (1996). Effect of ambient ozone exposure on lung function in farm workers. Am J
Respir Crit Care Med 154: 981-987.
Brauer. M; Brook. JR. (1997). Ozone personal exposures and health effects for selected groups residing in the
Fraser Valley. Atmos Environ 31: 2113-2121.
Braun-Fahrlander. C. h; Kunzli. N; Domenighetti. G: Carell. CF; Ackermann-Liebrich. U. (1994). Acute
effects of ambient ozone on respiratory function of Swiss schoolchildren after a 10-minute heavy exercise.
Pediatr Pulmonol 17: 169-177. http://dx.doi.org/10.1002/ppul.1950170306
Brook. RD; Brook. JR; Urch. B; Vincent. R; Raiagopalan. S: Silverman. F. (2002). Inhalation of fine
paniculate air pollution and ozone causes acute arterial vasoconstriction in healthy adults. Circulation 105:
1534-1536.
Brooks. EG. (2010). Correspondence from Dr. Brooks - Re: Clarifications in 2008 J Occup Environ Med
article. Brooks, EG.
Brown. JS: Bateson. TF; McDonnell. WE (2008). Effects of exposure to 0.06 ppm ozone on FEV1 in humans:
A secondary analysis of existing data. Environ Health Perspect 116: 1023-1026.
http://dx.doi.org/10.1289/ehp.11396
Brunekreef. B; Hoek. G: Breugelmans. O: Leentvaar. M. (1994). Respiratory effects of low-level
photochemical air pollution in amateur cyclists. Am J Respir Crit Care Med 150: 962-966.
Buadong. D; Jinsart. W: Funatagawa. I; Karita. K; Yano. E. (2009). Association between PM10 and O3 levels
and hospital visits for cardiovascular diseases in Bangkok, Thailand. J Epidemiol 19: 182-188.
http://dx.doi.org/10.2188/iea.JE20080047
Buckley. RD: Hackney. JD; Clark. K; Posin. C. (1975). Ozone and human blood. Arch Environ Health 30: 40-
43.
Burleson. GR; Keves. LL; Stutzman. JD. (1989). Immunosuppression of pulmonary natural killer activity by
exposure to ozone. Immunopharmacol Immunotoxicol 11: 715-735.
http://dx.doi.org/10.3109/08923978909005397
Burnett. R; Raizenne. M; Krewski. D. (1990). Acute health effects of transported air pollution: A study of
children attending a residential summer camp. Can J Stat 18: 367-373. http://dx.doi.org/10.2307/3315843
Burra. TA; Moineddin. R; Agha. MM; Glazier. RH. (2009). Social disadvantage, air pollution, and asthma
physician visits in Toronto, Canada. Environ Res 109: 567-574.
http://dx.doi.0rg/10.1016/i.envres.2009.03.004
Bush. ML; Asplund. PT; Miles. KA; Ben-Jebria. A; Ultman. JS. (1996). Longitudinal distribution of O3
absorption in the lung: gender differences and intersubject variability. J Appl Physiol 81: 1651-1657.
Cakmak. S: Dales. RE; Angelica Rubio. M; Blanco Vidal. C. (2011). The risk of dying on days of higher air
pollution among the socially disadvantaged elderly. Environ Res 111: 388-393.
http://dx.doi.0rg/10.1016/i.envres.2011.01.003
Cakmak. S: Dales. RE; Judek. S. (2006a). Do gender, education, and income modify the effect of air pollution
gases on cardiac disease? J Occup Environ Med 48: 89-94.
http://dx.doi.org/10.1097/01.jom.0000184878.11956.4b
Cakmak. S: Dales. RE; Judek. S. (2006b). Respiratory health effects of air pollution gases: Modification by
education and income. Arch Environ Occup Health 61: 5-10. http://dx.doi.Org/10.3200/AEOH.61.l.5-10
Calderon Guzman. D; Barragan Mejia. G: Hernandez Garcia. E; Juarez Olguin. H. (2006). Effect of nutritional
status and ozone exposure on some biomarkers of oxidative stress in rat brain regions. Nutr Cancer 55:
195-200. http://dx.doi.org/10.1207/sl5327914nc5502 11
6-269
-------
Calderon Guzman. D; Hernandez Mas. JL; Mejia. GB; Santamaria del Angel. D; Hernandez Garcia. E; Juarez
Olguin. H. (2005). Effect of nutritional status and ozone exposure on Na+/K+ATPpase and lipid
peroxidation in rat brain. Proc West Pharmacol Soc 48: 118-121.
Carpagnano. GE; Foschino Barbaro. MP; Cagnazzo. M; Di Gioia. G: Giliberti. T; Di Matteo. C: Resta. O.
(2005). Use of exhaled breath condensate in the study of airway inflammation after hypertonic saline
solution challenge. Chest 128: 3159-3166. http://dx.doi.Org/10.1378/chest.128.5.3159
Castillejos. M; Gold. PR; Damokosh. AI; Serrano. P; Allen. G: McDonnell. WF; Dockery. D; Velasco. SR;
Hernandez. M; Haves. C. (1995). Acute effects of ozone on the pulmonary function of exercising
schoolchildren from Mexico City. Am J Respir Crit Care Med 152: 1501-1507.
Chan. CC: Chuang. KJ; Chien. LC: Chen. WJ: Chang. WT. (2006). Urban air pollution and emergency
admissions for cerebrovascular diseases in Taipei, Taiwan. Eur Heart J 27: 1238-1244.
http://dx.doi.org/10.1093/eurheartj/ehi835
Chan. CC: Chuang. KJ: Su. TC: Lin. LY. (2005a). Association between nitrogen dioxide and heart rate
variability in a susceptible population. Eur J Cardiovasc Prev Rehabil 12: 580-586.
Chan. CC: Wu. TH. (2005). Effects of ambient ozone exposure on mail carriers' peak expiratory flow rates.
Environ Health Perspect 113: 735-738. http://dx.doi.org/10.1289/ehp.7636
Chang. CC: Tsai. SS: Ho. SC: Yang. CY. (2005). Air pollution and hospital admissions for cardiovascular
disease in Taipei, Taiwan. Environ Res 98: 114-119.
Chang. MM; Wu. R; Plopper. CG: Hyde. DM. (1998). IL-8 is one of the major chemokines produced by
monkey airway epithelium after ozone-induced injury. Am J Physiol 275: L524-L532.
Chen. C: Arjomandi. M; Balmes. J: Tager. I; N. H. (2007a). Effects of chronic and acute ozone exposure on
lipid peroxidation and antioxidant capacity in healthy young adults. Environ Health Perspect 115: 1732-
1737. http://dx.doi.org/10.1289/ehp.10294
Chen. JC: Schwartz. J. (2009). Neurobehavioral effects of ambient air pollution on cognitive performance in
US adults. Neurotoxicology 30: 231-239. http://dx.doi.Org/10.1016/i.neuro.2008.12.011
Chen. L; Jennison. BL; Yang. W: Omave. ST. (2000). Elementary school absenteeism and air pollution. Inhal
Toxicol 12: 997-1016. http://dx.doi.org/10.1080/08958370050164626
Chen. PC: Lai. YM; Chan. CC: Hwang. JS: Yang. CY; Wang. JD. (1999). Short-term effect of ozone on the
pulmonary function of children in primary school. Environ Health Perspect 107: 921-925.
http://dx.doi.org/10.1289/ehp.99107921
Chen. XQ; Yang. J: Hu. SP; Nie. HX; Mao. GY; Chen. HB. (2006b). Increased expression of CD86 and
reduced production of IL-12 and IL-10 by monocyte-derived dendritic cells from allergic asthmatics and
their effects on Thl- and Th2-type cytokine balance. Respiration 73: 34-40.
http://dx.doi.org/10.1159/000087457
Chhabra. SK; Yasir. A; Chaudhry. K; Shah. B. (2010). Effect of ozone on response to ovalbumin & its
modulation by vitamins C & E in sensitized guinea pigs. Indian J Med Res 132: 87-93.
Chimenti. L; Morici. G: Paterno. A; Bonanno. A; Vultaggio. M; Bellia. V: Bonsignore. MR. (2009).
Environmental conditions, air pollutants, and airway cells in runners: A longitudinal field study. J Sports
Sci 27: 925-935. http://dx.doi.org/10.1080/02640410902946493
Chiu. HF; Cheng. MH; Yang. CY. (2009). Air pollution and hospital admissions for pneumonia in a subtropical
city: Taipei, Taiwan. Inhal Toxicol 21: 32-37. http://dx.doi.org/10.1080/08958370802441198
Chiu. HF; Yang. CY. (2009). Air pollution and emergency room visits for arrhythmias: Are there potentially
sensitive groups? J Toxicol Environ Health A 72: 817-823. http://dx.doi.org/10.1080/15287390902800405
Choi. JH: Xu. OS: Park. SY: Kim. JH: Hwang. SS: Lee. KH: Lee. HJ: Hong. YC. (2007). Seasonal variation of
effect of air pollution on blood pressure. J Epidemiol Community Health 61: 314-318.
6-270
-------
Christian. PL; Chen. LL; Scannell. CH; Ferrando. RE; Welch. BS: Balmes. JR. (1998). Ozone-induced
inflammation is attenuated with multiday exposure. Am J Respir Crit Care Med 158: 532-537.
Chuang. GC: Yang. Z; Westbrook. DG: Pompilius. M; Ballinger. CA; White. RC: Krzywanski. DM;
Postlethwait. EM; Ballinger. SW. (2009). Pulmonary ozone exposure induces vascular dysfunction,
mitochondria! damage, and atherogenesis. Am J Physiol Lung Cell Mol Physiol 297: L209-L216.
http://dx.doi.org/10.1152/ajplung.00102.2009
Chuang. KJ; Chan. CC: Su. TC: Lee. CT; Tang. CS. (2007a). The effect of urban air pollution on
inflammation, oxidative stress, coagulation, and autonomic dysfunction in young adults. Am J Respir Crit
Care Med 176: 370-376. http://dx.doi.org/10.1164/rccm.200611-1627OC
Chuang. KJ: Chan. CC: Su. TC: Lin. LY; Lee. CT. (2007b). Associations between paniculate sulfate and
organic carbon exposures and heart rate variability in patients with or at risk for cardiovascular diseases. J
Occup Environ Med 49: 610-617. http://dx.doi.org/10.1097/JOM.Ob013e318058205b
Chuang. KJ: Yan. YH; Cheng. TJ. (2010). Effect of air pollution on blood pressure, blood lipids, and blood
sugar: A population-based approach. J Occup Environ Med 52: 258-262.
http://dx.doi.org/10.1097/JOM.Ob013e3181ceff7a
demons. GK; Garcia. JF. (1980). Changes in thyroid function after short-term ozone exposure in rats. J
Environ Pathol Toxicol 4: 359-369.
Cockcroft. DW: Davis. BE; Todd. DC: Smycniuk. AJ. (2005). Methacholine challenge: Comparison of two
methods. Chest 127: 839-844.
Coffin. PL; Blommer. EJ; Gardner. DE; Holzman. R. (1967). Effect of air pollution on alteration of
susceptibility to pulmonary infection (pp. 71-80). (AMRL-TR-67-200). Cincinnati, OH: Aerospace Medical
Research Laboratories.
Coffin. PL; Gardner. DE. (1972). Interaction of biological agents and chemical air pollutants. Ann Occup Hyg
15: 219-234.
Cohen. MD; Sisco. M; Baker. K; Li. Y; Lawrence. D; Van Loveren. H; Zelikoff. JT; Schlesinger. RB. (2002).
Effects of inhaled ozone on pulmonary immune cells critical to antibacterial responses in situ. Inhal Toxicol
14: 599-619. http://dx.doi.org/10.1080/08958370290084520
Cole. MP; Freeman. BA. (2009). Promotion of cardiovascular disease by exposure to the air pollutant ozone
[Review]. Am J Physiol Lung Cell Mol Physiol 297: L209-L216.
http://dx.doi.org/10.1152/ajplung.00187.2009
Colin-Barenque. L; Dorado-Martinez. C: Rivas-Arancibia. S: Avila-Costa. MR; Fortoul. TI. (2005).
Morphological recovery of the granule cells from the olfactory bulb after the cessation of acute ozone
exposure. Int JNeurosci 115: 411-421. http://dx.doi.org/10.1080/00207450590521028
Corradi. M; Alinovi. R; Goldoni. M; Vettori. M; Folesani. G: Mozzoni. P; Cavazzini. S: Bergamaschi. E;
Rossi. L; Mutti. A. (2002). Biomarkers of oxidative stress after controlled human exposure to ozone.
Toxicol Lett 134:219-225.
Corradi. M; Folesani. G: Andreoli. R: Manini. P; Bodini. A; Piacentini. G: Carraro. S: Zanconato. S: Baraldi.
K (2003). Aldehydes and glutathione in exhaled breath condensate of children with asthma exacerbation.
Am J Respir Crit Care Med 167: 395-399. http://dx.doi.org/10.1164/rccm.200206-507OC
Cremillieux. Y; Servais. S: Berthezene. Y; Dupuich. D; Boussouar. A; Stupar. V; Pequignot. JM. (2008).
Effects of ozone exposure in rat lungs investigated with hyperpolarized 3He MRI. J Magn Reson Imaging
27: 771-776. http://dx.doi.org/10.1002/imri.21216
Curriero. FC: Heiner. KS: Samet. JM; Zeger. SL; Strug. L; Patz. JA. (2002). Temperature and mortality in 11
cities of the eastern United States. Am J Epidemiol 155: 80-87. http://dx.doi.Org/10.1093/aie/155.l.80
Dales. R: Chen. L; Frescura. AM; Liu. L; Villeneuve. PJ. (2009). Acute effects of outdoor air pollution on
forced expiratory volume in 1 s: A panel study of schoolchildren with asthma. Eur Respir J 34: 316-323.
http://dx.doi.org/10.1183/09031936.00138908
6-271
-------
Dales. RE; Cakmak. S: Doiron. MS. (2006). Gaseous air pollutants and hospitalization for respiratory disease
in the neonatal period. Environ Health Perspect 114: 1751-1754. http://dx.doi.org/10.1289/ehp.9044
Damera. G: Jester William. F; Jiang. M; Zhao. H; Fogle Homer. W: Mittelman. M; Haczku. A; Murphy. E;
Parikh. I; Panettieri Reynold. A. (2010). Inhibition of myristoylated alanine-rich C kinase substrate
(MARCKS) protein inhibits ozone-induced airway neutrophilia and inflammation. Exp Lung Res 36: 75-
84. http://dx.doi.org/10.3109/01902140903131200
Darrow. LA; Klein. M; Sarnat JA; Mulholland. JA; Strickland. MJ; Sarnat SE; Russell. AG: Tolbert. PE.
(2011 a). The use of alternative pollutant metrics in time-series studies of ambient air pollution and
respiratory emergency department visits. J Expo Sci Environ Epidemiol 21: 10-19.
http://dx.doi.org/10.1038/ies.2009.49
De Pablo. F; Lopez. A; Soriano. LR; Tomas. C: Diego. L; Gonzalez. M; Barrueco. M. (2006). Relationships of
daily mortality and hospital admissions to air pollution in Castilla-Leon, Spain. Atmosfera 19: 23-39.
Delfino. RJ; Gillen. PL; Tjoa. T; Staimer. N; Polidori. A; Arhami. M; Sioutas. C: Longhurst. J. (2011).
Electrocardiographic ST-segment depression and exposure to trafficrelated aerosols in elderly subjects with
coronary artery disease. Environ Health Perspect 119: 196-202. http://dx.doi.org/10.1289/ehp.1002372
Delfino. RJ: Gone. H; Linn. WS: Pellizzari. ED; Hu. Y. (2003). Asthma symptoms in Hispanic children and
daily ambient exposures to toxic and criteria air pollutants. Environ Health Perspect 111: 647-656.
http://dx.doi.org/10.1289/ehp.5992
Delfino. RJ: Quintana. PJE; Floro. J: Gastanaga. VM; Samimi. BS: Kleinman. MT; Liu. LJS: Bufalino. C: Wu.
CF; McLaren. CE. (2004). Association of FEV1 in asthmatic children with personal and
microenvironmental exposure to airborne paniculate matter. Environ Health Perspect 112: 932-941.
Delfino. RJ: Staimer. N; Tjoa. T; Arhami. M; Polidori. A; Gillen. PL; George. SC: Shafer. MM; Schauer. JJ;
Sioutas. C. (2010a). Associations of primary and secondary organic aerosols with airway and systemic
inflammation in an elderly panel cohort. Epidemiology 21: 892-902.
http://dx.doi.org/10.1097/EDE.Ob013e3181f20e6c
Delfino. RJ: Tjoa. T; Gillen. PL; Staimer. N; Polidori. A; Arhami. M; Jamner. L; Sioutas. C: Longhurst. J.
(2010b). Traffic-related air pollution and blood pressure in elderly subjects with coronary artery disease.
Epidemiology 21: 396-404. http://dx.doi.org/10.1097/EDE.Ob013e3181d5el9b
Delfino. RJ: Zeiger. RS: Seltzer. JM; Street. DH; Matteucci. RM; Anderson. PR; Koutrakis. P. (1997).
The effect of outdoor fungal spore concentrations on daily asthma severity. Environ Health Perspect 105:
622-635.
DeLucia. AJ; Adams. WC. (1977). Effects of O3 inhalation during exercise on pulmonary function and blood
biochemistry. JAppl Physiol 43: 75-81.
Dennekamp. M; Akram. M; Abramson. MJ: Tonkin. A; Sim. MR; Fridman. M; Erbas. B. (2010). Outdoor air
pollution as a trigger for out-of-hospital cardiac arrests. Epidemiology 21: 494-500.
http://dx.doi.org/10.1097/EDE.Ob013e3181e093db
Depuvdt. P; Joos. GF; Pauwels. RA. (1999). Ambient ozone concentrations induce airway hyperresponsiveness
in some rat strains. Eur Respir J 14: 125-131.
Devlin. RB; Duncan. KE; Jardim. M; Schmitt. MT; Rappold. AG: Diaz-Sanchez. D. (2012). Controlled
exposure of healthy young volunteers to ozone causes cardiovascular effects. Circulation 126: 104-111.
http://dx.doi.org/10.1161/CIRCULATIONAHA.112.094359
Devlin. RB; Folinsbee. LJ; Biscardi. F; Hatch. G: Becker. S: Madden. MC: Robbins. M; Koren. HS. (1997).
Inflammation and cell damage induced by repeated exposure of humans to ozone. Inhal Toxicol 9: 211-235.
Devlin. RB: Mcdonnell. WF; Becker. S: Madden. MC: Mcgee. MP; Perez. R; Hatch. G: House. DE; Koren.
HS. (1996). Time-dependent changes of inflammatory mediators in the lungs of humans exposed to 0.4
ppm ozone for 2 hr: A comparison of mediators found in bronchoalveolar lavage fluid 1 and 18 hr after
exposure. Toxicol Appl Pharmacol 138: 176-185. http://dx.doi.org/10.1006/taap.1996.0111
6-272
-------
Devlin. RB: Mcdonnell. WF; Mann. R; Becker. S: House. DE; Schreinemachers. D; Koren. HS. (1991).
Exposure of humans to ambient levels of ozone for 6.6 hours causes cellular and biochemical changes in
the lung. Am J Respir Cell Mol Biol 4: 72-81.
Dimeo. MJ; Glenn. MG: Holtzman. MJ; Sheller. JR; Nadel. JA; Boushev. HA. (1981). Threshold concentration
of ozone causing an increase in bronchial reactivity in humans and adaptation with repeated exposures. Am
Rev Respir Dis 124: 245-248.
Dimitriadis. VK. (1992). Carbohydrate cytochemistry of bonnet monkey (Macaca radiata) nasal epithelium:
Response to ambient levels of ozone. Histol Histopathol 7: 479-488.
Dockerv. DW: Luttmann-Gibson. H; Rich. DQ; Link. MS: Mittleman. MA; Gold. PR; Koutrakis. P; Schwartz.
JD; Verrier. RL. (2005). Association of air pollution with increased incidence of ventricular
tachyarrhythmias recorded by implanted cardioverter defibrillators. Environ Health Perspect 113: 670-674.
Dohm. MR; Mautz. WJ: Andrade. JA; Gellert. KS: Salas-Ferguson. LJ; Nicolaisen. N; Fujie. N. (2005). Effects
of ozone exposure on nonspecific phagocytic capacity of pulmonary macrophages from an amphibian,
Bufo marinus. Environ Toxicol Chem 24: 205-210.
Dorado-Martinez. C: Parades-Carbaial. C: Mascher. D; Borgonio-Perez. G: Rivas-Arancibia. S. (2001). Effects
of different ozone doses on memory, motor activity and lipid peroxidation levels, in rats. Int J Neurosci
108: 149-161.
Driscoll. KE; Vollmuth. TA; Schlesinger. RB. (1987). Acute and subchronic ozone inhalation in the rabbit:
Response of alveolar macrophages. J Toxicol Environ Health 21: 27-43.
http://dx.doi.org/10.1080/15287398709531000
Drvden. DM; Spooner. CH; Stickland. MK; Vandermeer. B; Tjosvold. L; Bialy. L; Wong. K; Rowe. BH.
(2010). Exercise-induced bronchoconstriction and asthma. (AHRQ Publication No. 10-E001). Rockville,
MD: Agency for Healthcare Research and Quality.
Dungworth. PL. (1976). Short-term effects of ozone on lungs of rats, mice, and monkeys. Environ Health
Perspect 13: 179.
Dungworth. PL; Castleman. WL; Chow. CK; Mellick. PW: Mustafa. MG: Tarkington. B; Tyler. WS. (1975).
Effect of ambient levels of ozone on monkeys. Fed Proc 34: 1670-1674.
Duramad. P; Tager. IB; Holland. NT. (2007). Cytokines and other immunological biomarkers in children's
environmental health studies [Review]. Toxicol Lett 172: 48-59.
http://dx.doi.0rg/10.1016/i.toxlet.2007.05.017
Eiswerth. ME; Shaw. WD; Yen. ST. (2005). Impacts of ozone on the activities of asthmatics: Revisiting the
data. J Environ Manage 77: 56-63. http://dx.doi.Org/10.1016/i.jenvman.2005.02.010
Escalante-Membrillo. C: Gonzalez-Maciel. A; Reynoso-Robles. R; Gonzalez-Pina. R. (2005). Brain
thiobarbituric acid-reactive substances in rats after short periods of ozone exposure. Environ Res 99: 68-71.
http://dx.doi.0rg/10.1016/i.envres.2005.02.006
Escamilla-Nunez. MC: Barraza-Villarreal. A; Hernandez-Cadena. L; Moreno-Macias. H; Ramirez-Aguilar. M;
Sienra-Monge. JJ: Cortez-Lugo. M; Texcalac. JL; del Rio-Navarro. B; Romieu. I. (2008). Traffic-related air
pollution and respiratory symptoms among asthmatic children, resident in Mexico City: The EVA cohort
study. Respir Res 9: 74. http://dx.doi.org/10.1186/1465-9921-9-74
Fakhri. AA; Ilic. LM; Wellenius. GA; Urch. B; Silverman. F; Gold. PR; Mittleman. MA. (2009). Autonomic
effects of controlled fine paniculate exposure in young healthy adults: Effect modification by ozone.
Environ Health Perspect 117: 1287-1292. http://dx.doi.org/10.1289/ehp.0900541
Farrai. AK; Boykin. E; Ledbetter. A; Andrews. D: Gavett. SH. (2010). Increased lung resistance after diesel
particulate and ozone co-exposure not associated with enhanced lung inflammation in allergic mice. Inhal
Toxicol 22: 33-41. http://dx.doi.org/10.3109/08958370902862434
Feng. R; He. W: Ochi. H; Castranova. V. (2006). Ozone exposure impairs antigen-specific immunity but
activates IL-7-induced proliferation of CD4-CD8- thymocytes in BALB/c mice. J Toxicol Environ Health
A69: 1511-1526. http://dx.doi.org/10.1080/15287390500468696
6-273
-------
Feo Brito. F; Mur Gimeno. P; Martinez. C: Tobias. A; Suarez. L; Guerra. F; Borja. JM; Alonso. AM. (2007).
Air pollution and seasonal asthma during the pollen season: A cohort study in Puertollano and Ciudad Real
(Spain). Allergy 62: 1152-1157. http://dx.doi.Org/10.llll/i.1398-9995.2007.01438.x
Ferdinands. JM; Crawford. CA; Greenwald. R; Van Sickle. D; Hunter. E; Teague. WG. (2008). Breath
acidification in adolescent runners exposed to atmospheric pollution: A prospective, repeated measures
observational study. Environ Health Global Access Sci Source 7: 11. http://dx.doi.org/10.1186/1476-069X-
7-10
Folinsbee. LJ; Bedi. JF; Horvath. SM. (1980). Respiratory responses in humans repeatedly exposed to low
concentrations of ozone. Am Rev Respir Dis 121: 431-439.
Folinsbee. LJ: Bedi. JF; Horvath. SM. (1984). Pulmonary function changes after 1 h continuous heavy exercise
in 0.21 ppm ozone. J Appl Physiol 57: 984-988.
Folinsbee. LJ: Devlin. RB; Abdul-Salaam. S: Koren. HS. (1993). Repeated severe ozone exposure causes
depressed baseline spirometry [Abstract]. Am Rev Respir Dis 147: A638.
Folinsbee. LJ: Drinkwater. BL; Bedi. JF: Horvath. SM. (1978). The influence of exercise on the pulmonary
function changes due to exposure to low concentrations of ozone. In LJ Folinsbee; JA Wagner; JF Borgia;
BL Drinkwater; JA Gliner; JF Bedi (Eds.), Environmental stress: Individual human adaptations (pp. 125-
145). New York, NY: Academic Press.
Folinsbee. LJ; Hazucha. MJ. (1989). Persistence of ozone-induced changes in lung function and airway
responsiveness. In Atmospheric ozone research and its policy implications. Amsterdam, The Netherlands:
Elsevier.
Folinsbee. LJ; Hazucha. MJ. (2000). Time course of response to ozone exposure in healthy adult females. Inhal
Toxicol 12: 151-167.
Folinsbee. LJ; Horstman. DH; Kehrl. HR; Harder. S; Abdul-Salaam. S; Ives. PJ. (1994). Respiratory responses
to repeated prolonged exposure to 0.12 ppm ozone. Am J Respir Crit Care Med 149: 98-105.
Folinsbee. LJ; McDonnell. WF; Horstman. DH. (1988). Pulmonary function and symptom responses after 6.6-
hour exposure to 0.12 ppm ozone with moderate exercise. J Air Waste Manag Assoc 38: 28-35.
Folinsbee. LJ; Silverman. F; Shephard. RJ. (1977). Decrease of maximum work performance following ozone
exposure. JAppl Physiol 42: 531-536.
Folnsbee. LJ; Devlin. RB; Robbins. MK; Biscardi. FH; Abdul-Salaam. S; Koren. HS. (1995). Repeated
exposure of humans to ozone: I. Pulmonary function and symptom responses. Available online
Fortino. V; Maioli. E; Torricelli. C; Davis. P; Valacchi. G. (2007). Cutaneous MMPs are differently modulated
by environmental stressors in old and young mice. Toxicol Lett 173: 73-79.
http://dx.doi.0rg/10.1016/i.toxlet.2007.06.004
Foster. WM; Silver. JA; Groth. ML. (1993). Exposure to ozone alters regional function and particle dosimetry
in the human lung. JAppl Physiol 75: 1938-1945.
Foster. WM; Stetkiewicz. PT. (1996). Regional clearance of solute from the respiratory epithelia: 18-20 h
postexposure to ozone. JAppl Physiol 81: 1143-1149.
Foster. WM; Weinmann. GG; Menkes. E; Macri. K. (1997). Acute exposure of humans to ozone impairs small
airway function. Ann Occup Hyg 1: 659-666.
Foster. WM; Wills-Karp. M; Tankerslev. CG; Chen. X; Paquette. NC. (1996). Bloodborne markers in humans
during multiday exposure to ozone. JAppl Physiol 81: 794-800.
Fox. SD; Adams. WC; Brookes. KA; Laslev. BL. (1993). Enhanced response to ozone exposure during the
follicular phase of the menstrual cycle. Environ Health Perspect 101: 242-244.
Foxcroft. WJ; Adams. WC. (1986). Effects of ozone exposure on four consecutive days on work performance
and VO2max. JAppl Physiol 61: 960-966.
6-274
-------
Frampton. MW: Morrow. PE; Torres. A; Cox. C: Voter. KZ; Utell. MJ; Gibb. FR; Speers. DM. (1997a). Ozone
responsiveness in smokers and nonsmokers. Am J Respir Crit Care Med 155: 116-121.
Frank. R; Liu. MC: Spannhake. EW: Mlynarek. S: Macri. K; Weinmann. GG. (2001). Repetitive ozone
exposure of young adults: Evidence of persistent small airway dysfunction. Am J Respir Crit Care Med
164: 1253-1260.
Franklin. M; Schwartz. J. (2008). The impact of secondary particles on the association between ambient ozone
and mortality. Environ Health Perspect 116: 453-458. http://dx.doi.org/10.1289/ehp.10777
Franze. T; Weller. MG: Niessner. R; Poschl. U. (2005). Protein nitration by polluted air. Environ Sci Technol
39: 1673-1678. http://dx.doi.org/10.1021/es0488737
Friedman. M; Gallo. JM; Nichols. HP; Bromberg. PA. (1983). Changes in inert gas rebreathing parameters
after ozone exposure in dogs. Am Rev Respir Dis 128: 851-856.
Fung. KY; Luginaah. I; Gorev. KM; Webster. G. (2005). Air pollution and daily hospital admissions for
cardiovascular diseases in Windsor, Ontario. Can J Public Health 96: 29-33.
Gao. X; Raghavamenon. AC: D'Auvergne. O: Uppu. RM. (2009b). Cholesterol secoaldehyde induces
apoptosis in J774 macrophages via mitochondria! pathway but not involving reactive oxygen species as
mediators. Biochem Biophys Res Commun 389: 382-387. http://dx.doi.Org/10.1016/i.bbrc.2009.09.005
Garantziotis. S: Li. Z; Potts. EN; Lindsev. JY; Stober. VP; Polosukhin. W; Blackwell. TS: Schwartz. DA;
Foster. WM; Hollingsworth. JW. (2010). TLR4 is necessary for hyaluronan-mediated airway
hyperresponsiveness after ozone inhalation. Am J Respir Crit Care Med 181: 666-675.
http://dx.doi.org/10.1164/rccm.200903-0381OC
Gent. JF; Triche. EW: Holford. TR; Belanger. K; Bracken. MB; Beckett. WS: Leaderer. BP (2003).
Association of low-level ozone and fine particles with respiratory symptoms in children with asthma.
JAMA 290: 1859-1867. http://dx.doi.org/10.1001/iama.290.14.1859
Gershwin. LJ; Osebold. JW: Zee. YC. (1981). Immunoglobulin E-containing cells in mouse lung following
allergen inhalation and ozone exposure. International Arch Allergy Appl Immunol 65: 266-277.
Gielen. MH; Van Per Zee. SC: Van Wijnen. JH; Van Steen. CJ: Brunekreef. B. (1997). Acute effects of
summer air pollution on respiratory health of asthmatic children. Am J Respir Crit Care Med 155: 2105-
2108.
Gilliland. FD; Berhane. K; Rappaport. EB; Thomas. DC: Avol. E; Gauderman. WJ: London. SJ: Margolis. HG:
McConnell. R: Islam. KT; Peters. JM. (2001). The effects of ambient air pollution on school absenteeism
due to respiratory illnesses. Epidemiology 12: 43-54.
Gilmour. MI; Jakab. GJ. (1991). Modulation of immune function in mice exposed to 0.8 ppm ozone. Inhal
Toxicol 3: 293-308.
Girardot. SP; Ryan. PB; Smith. SM; Davis. WT; Hamilton. CB; Obenour. RA; Renfro. JR; Tromatore. KA;
Reed. GD. (2006). Ozone and PM2.5 exposure and acute pulmonary health effects: A study of hikers in the
Great Smoky Mountains National Park. Environ Health Perspect 113: 612-617.
http://dx.doi.org/10.1289/ehp.8637
Gold. PR; Damokosh. AI; III. PC: Dockery. DW: McDonnell. WF; Serrano. P; Retama. A; Castillejos. M.
(1999). Paniculate and ozone pollutant effects on the respiratory function of children in southwest Mexico
City. Epidemiology 10: 8-16.
Goldberg. MS: Giannetti. N; Burnett. RT; Mayo. NE; Valois. MF; Brophy. JM. (2008). A panel study in
congestive heart failure to estimate the short-term effects from personal factors and environmental
conditions on oxygen saturation and pulse rate. Occup Environ Med 65: 659-666.
http://dx.doi.org/10.1136/oem.2007.034934
Gong. H. Jr; Bradley. PW: Simmons. MS: Tashkin. DP. (1986). Impaired exercise performance and pulmonary
function in elite cyclists during low-level ozone exposure in a hot environment. Am J Respir Crit Care Med
134: 726-733.
6-275
-------
Gong. H. Jr; McManus. MS: Linn. WS. (1997a). Attenuated response to repeated daily ozone exposures in
asthmatic subjects. Arch Environ Occup Health 52: 34-41.
Gong. H. Jr; Shamoo. DA; Anderson. KR: Linn. WS. (1997b). Responses of older men with and without
chronic obstructive pulmonary disease to prolonged ozone exposure. Arch Environ Occup Health 52: 18-
25.
Gong. H. Jr; Wong. R; Sarma. RJ: Linn. WS: Sullivan. ED; Shamoo. DA; Anderson. KR: Prasad. SB. (1998).
Cardiovascular effects of ozone exposure in human volunteers. Am J Respir Crit Care Med 158: 538-546.
Gonzalez-Pina. R; Escalante-Membrillo. C: Alfaro-Rodriguez. A; Gonzalez-Maciel. A. (2008). Prenatal
exposure to ozone disrupts cerebellar monoamine contents in newborn rats. Neurochem Res 33: 912-918.
http://dx.doi.org/10.1007/sll064-007-9534-3
Graham. JA; Menzel. DB; Miller. FJ; Illing. JW: Gardner. DE. (1981). Influence of ozone on pentobarbital-
induced sleeping time in mice, rats, and hamsters. Toxicol Appl Pharmacol 61: 64-73.
http://dx.doi.org/10.1016/0041-008X(81)90008-9
Grvparis. A; Forsberg. B; Katsouvanni. K; Analitis. A; Touloumi. G: Schwartz. J; Samoli. E; Medina. S:
Anderson. HR; Niciu. EM; Wichmann. HE; Kriz. B; Kosnik. M; Skorkovsky. J; Vonk. JM; Dortbudak. Z.
(2004). Acute effects of ozone on mortality from the "Air pollution and health: A European approach"
project. Am J Respir Crit Care Med 170: 1080-1087. http://dx.doi.org/10.1164/rccm.200403-333OC
Guerrero. AL; Dorado-Martinez. C: Rodriguez. A; Pedroza-Rios. K; Borgonio-Perez. G: Rivas-Arancibia. S.
(1999). Effects of vitamin E on ozone-induced memory deficits and lipid peroxidation in rats. Neuroreport
10: 1689-1692.
Guevara-Guzman. R; Arriaga. V: Kendrick. KM; Bernal. C: Vega. X; Mercado-Gomez. OF; Rivas-Arancibia.
S.. (2009). Estradiol prevents ozone-induced increases in brain lipid peroxidation and impaired social
recognition memory in female rats. Neuroscience 159: 940-950.
http://dx.doi.0rg/10.1016/i.neuroscience.2009.01.047
Hackney. JD; Linn. WS: Buckley. RD; Hislop. HJ. (1976). Studies in adaptation to ambient oxidant air
pollution: effects of ozone exposure in Los Angeles residents vs new arrivals. Environ Health Perspect 18:
141-146.
Hackney. JD: Linn. WS: Karuza. SK; Buckley. RD: Law. DC: Bates. DV: Hazucha. M; Pengelly. LD;
Silverman. F. (1977a). Effects of ozone exposure in Canadians and southern Californians: evidence for
adaptation? Arch Environ Occup Health 32: 110-116.
Hackney. JD: Linn. WS: Mohler. JG: Pedersen. EE; Breisacher. P: Russo. A. (1975). Experimental studies on
human health effects of air pollutants: II. Four-hour exposure to ozone alone and in combination with other
pollutant gases. Arch Environ Occup Health 30: 379-384.
Halonen. JI; Lanki. T; Tiittanen. P: Niemi. JV: Loh. M; Pekkanen. J. (2009). Ozone and cause-specific
cardiorespiratory morbidity and mortality. J Epidemiol Community Health 64: 814-820.
http://dx.doi.org/10.1136/iech.2009.087106
Hamade. AK; Misra. V: Rabold. R: Tankerslev. CG. (2010). Age-related changes in cardiac and respiratory
adaptation to acute ozone and carbon black exposures: Interstrain variation in mice. Inhal Toxicol 22: 84-
94. http://dx.doi.org/10.3109/08958378.2010.503974
Hamade. AK; Rabold. R: Tankerslev. CG. (2008). Adverse cardiovascular effects with acute paniculate matter
and ozone exposures: Interstrain variation in mice. Environ Health Perspect 116: 1033-1039.
http://dx.doi.org/10.1289/ehp.10689
Hamade. AK: Tankerslev. CG. (2009). Interstrain variation in cardiac and respiratory adaptation to repeated
ozone and paniculate matter exposures. Am J Physiol Regul Integr Comp Physiol 296: R1202-R1215.
http://dx.doi.org/10.1152/aipregu.90808.2008
Hampel. R: Breitner. S: Zareba. W: Kraus. U: Pitz. M; Geruschkat. U: Belcredi. P: Peters. A; Schneider. A;
Group. ftCHRitRoAKS. (In Press) Immediate ozone effects on heart rate and repolarisation parameters in
potentially susceptible individuals. Occup Environ Med. http://dx.doi.org/10.1136/oemed-2011-100179
6-276
-------
Han. SG: Andrews. R; Gairola. CG: Bhalla. DK. (2008). Acute pulmonary effects of combined exposure to
carbon nanotubes and ozone in mice. Inhal Toxicol 20: 391-398.
http://dx.doi.org/10.1080/08958370801904014
Harkema. JR: Plopper. CG: Hyde. DM; St George. JA; Dungworth. PL. (1987a). Effects of an ambient level of
ozone on primate nasal epithelial mucosubstances: quantitative histochemistry. Am J Pathol 127: 90-96.
Harkema. JR: Plopper. CG: Hyde. DM; St George. JA; Wilson. DW: Dungworth. PL. (1987b). Response of
the macaque nasal epithelium to ambient levels of ozone: A morphologic and morphometric study of the
transitional and respiratory epithelium. Am J Pathol 128: 29-44.
Harkema. JR: Plopper. CG: Hyde. PM; St George. JA: Wilson. PW: Pungworth. PL. (1993). Response of
macaque bronchiolar epithelium to ambient concentrations of ozone. Am J Pathol 143: 857-866.
Hatch. GE; Slade. R; Harris. LP; Mcdonnell. WF; Pevlin. RB; Koren. HS: Costa. PL; Mckee. J. (1994).
Ozone dose and effect in humans and rats: A comparison using oxygen-18 labeling and bronchoalveolar
lavage. Am J Respir Crit Care Med 150: 676-683.
Hazucha. MJ; Folinsbee. LJ; Bromberg. PA. (2003). Pistribution and reproducibility of spirometric response to
ozone by gender and age. J Appl Physiol 95: 1917-1925.
Hazucha. MJ: Folinsbee. LJ: Seal. E. Jr. (1992). Effects of steady-state and variable ozone concentration
profiles on pulmonary function. Am J Respir Crit Care Med 146: 1487-1493.
Hazucha. MJ: Lefohn. AS. (2007). Nonlinearity in human health response to ozone: Experimental laboratory
considerations. Atmos Environ 41: 4559-4570. http://dx.doi.Org/10.1016/i.atmosenv.2007.03.052
HEI (Health Effects Institute). (2003). Revised analyses of time-series studies of air pollution and health:
Revised analyses of the National Morbidity, Mortality, and Air Pollution Study (NMMAPS), Part II.
Cambridge, MA. http://pubs.healtheffects.org/view.php?id=4
Heidenfelder. BL; Reif. PM; Harkema. J. R.: Cohen Hubal. EA; Hudgens. EE; Bramble. LA; Wagner. JG:
Morishita. M; Keeler. GJ: Edwards. SW: Gallagher. JE. (2009). Comparative microarray analysis and
pulmonary changes in brown Norway rats exposed to ovalbumin and concentrated air particulates. Toxicol
Sci 108: 207-221. http://dx.doi.org/10.1093/toxsci/kfp005
Hemmingsen. A; Fryer. AA; Hepple. M; Strange. RC: Spiteri. MA. (2001). Simultaneous identification of
GSTP1 Ilel05->Vall05 and Alall4->Valll4 substitutions using an amplification refractory mutation
system polymerase chain reaction assay: Studies in patients with asthma. Respir Res 2: 255-260.
http://dx.doi.org/10.1186/rr64
Henderson. FW: Pubovi. EJ; Harder. S: Seal. E. Jr; Graham. P. (1988). Experimental rhinovirus infection in
human volunteers exposed to ozone. Am J Respir Crit Care Med 137: 1124-1128.
Henrotin. JB; Besancenot. JP; Bejot. Y; Giroud. M. (2007). Short-term effects of ozone air pollution on
ischaemic stroke occurrence: A case-crossover analysis from a 10-year population-based study in Pijon,
France. Occup Environ Med 64: 439-445. http://dx.doi.org/10.1136/oem.2006.029306
Henrotin. JB; Zeller. M; Lorgis. L; Cottin. Y; Giroud. M; Bejot. Y. (2010). Evidence of the role of short-term
exposure to ozone on ischaemic cerebral and cardiac events: The Pijon Vascular Project (PIVA). Heart 96:
1990-1996. http://dx.doi.org/10.1136/hrt.2010.200337
Hernandez-Cadena. L; Holguin. F; Barraza-Villarreal. A; Pel Rio-Navarro. BE; Sienra-Monge. JJ: Romieu. I.
(2009). Increased levels of outdoor air pollutants are associated with reduced bronchodilation in children
with asthma. Chest 136: 1529-1536. http://dx.doi.org/10.1378/chest.08-1463
Hernandez. ML; Lav. JC: Harris. B; Esther. CR; Brickev. WJ: Bromberg. PA; Piaz-Sanchez. P; Pevlin. RB;
Kleeberger. SR; Alexis. NE; Peden. PB. (2010). Atopic asthmatic subjects but not atopic subjects without
asthma have enhanced inflammatory response to ozone. J Allergy Clin Immunol 126: 537-544.
http://dx.doi.0rg/10.1016/i.iaci.2010.06.043
6-277
-------
Hicks. A; Goodnow. R. Jr; Cavallo. G: Tannu. SA; Ventre. JD; Lavelle. D; Lora. JM; Satjawatcharaphong. J;
Brovarnev. M; Dabbagh. K; Tare. NS: Oh. H; Lamb. M; Sidduri. A; Dominique. R; Qiao. Q: Lou. JP;
Gillespie. P; Fotouhi. N; Kowalczyk. A; Kurylko. G: Hamid. R; Wright. MB; Pamidimukkala. A; Egan. T;
Gubler. U; Hoffman. AF; Wei. X; Li. YL; O'Neil. J: Marcano. R; Pozzani. K; Molinaro. T; Santiago. J:
Singer. L; Hargaden. M; Moore. D; Catala. AR; Qiao. LC: Benson. J; March. T; Venkat. R: Mancebo. H;
Renzetti. LM. (2010a). Effects of LTB4 receptor antagonism on pulmonary inflammation in rodents and
non-human primates. Prostaglandins Other Lipid Mediat 92: 33-43.
http://dx.doi.0rg/10.1016/i.prostaglandins.2010.02.003
Hicks. A; Kourteva. G: Hilton. H; Li. H; Lin. T; Liao. W: Li. Y; Wei. X; March. T; Benson. J: Renzetti. L.
(201 Ob). Cellular and molecular characterization of ozone-induced pulmonary inflammation in the
Cynomolgus monkey. Inflammation 33: 144-156. http://dx.doi.org/10.1007/sl0753-009-9168-5
Higgins. ITT: D'Arcv. JB; Gibbons. PI; Avol. EL; Gross. KB. (1990). Effect of exposures to ambient ozone on
ventilatory lung function in children. Am J Respir Crit Care Med 141: 1136-1146.
Hiltermann. JTN; Lapperre. TS: Van Bree. L; Steerenberg. PA; Brahim. JJ; Sont. JK; Sterk. PJ; Hiemstra. PS:
Stolk. J. (1999). Ozone-induced inflammation assessed in sputum and bronchial lavage fluid from
asthmatics: A new noninvasive tool in epidemiologic studies on air pollution and asthma. Free Radic Biol
Med 27: 1448-1454.
Hiltermann. TJN: Peters. EA; Alberts. B; Kwikkers. K; Borggreven. PA; Hiemstra. PS: Dijkman. JH; van Bree.
LA; Stolk. J. (1998). Ozone-induced airway hyperresponsiveness in patients with asthma: Role of
neutrophil-derived serine proteinases. Free Radic Biol Med 24: 952-958.
Hiltermann. TJN: Stolk. J: Hiemstra. PS: Fokkens. PHB; Rombout. PJA; Sont. JK; Sterk. PJ: Dijkman. JH.
(1995). Effect of ozone exposure on maximal airway narrowing in non-asthmatic and asthmatic subjects.
Clin Sci (Lond) 89: 619-624.
Hinwood. AL; De Klerk. N; Rodriguez. C: Jacobv. P; Runnion. T; Rye. P; Landau. L; Murray. F; Feldwick. M;
Spickett. J. (2006). The relationship between changes in daily air pollution and hospitalizations in Perth,
Australia 1992-1998: A case-crossover study. Int J Environ Health Res 16: 27-46.
http://dx.doi.org/10.1080/09603120500397680
Hoek. G: Brunekreef. B. (1995). Effect of photochemical air pollution on acute respiratory symptoms in
children. Am J Respir Crit Care Med 151: 27-32.
Hoek. G: Brunekreef. B; Kosterink. P; Van den Berg. R; Hofschreuder. P. (1993). Effect of ambient ozone on
peak expiratory flow of exercising children in the Netherlands. Arch Environ Occup Health 48: 27-32.
http://dx.doi.org/10.1080/00039896.1993.9938390
Hollingsworth. JW: Cook. DN; Brass. DM; Walker. JKL; Morgan. PL; Foster. WM; Schwartz. DA. (2004).
The role of Toll-like receptor 4 in environmental airway injury in mice. Am J Respir Crit Care Med 170:
126-132. http://dx.doi.org/10.1164/rccm.200311-1499OC
Hollingsworth. JW: Free. ME; Li. Z; Andrews. LN; Nakano. H; Cook. DN. (2010). Ozone activates pulmonary
dendritic cells and promotes allergic sensitization through a Toll-like receptor 4-dependent mechanism
[Letter]. J Allergy Clin Immunol 125: 1167-1170. http://dx.doi.Org/10.1016/i.iaci.2010.03.001
Holz. O; Jorres. RA; Timm. P; Mucke. M; Richter. K; Koschyk. S: Magnussen. H. (1999). Ozone-induced
airway inflammatory changes differ between individuals and are reproducible. Am J Respir Crit Care Med
159: 776-784.
Holz. O; Mucke. M; Paasch. K; Bohme. S: Timm. P; Richter. K; Magnussen. H; Jorres. RA. (2002). Repeated
ozone exposures enhance bronchial allergen responses in subjects with rhinitis or asthma. Clin Exp Allergy
32: 681-689.
Holz. O; Tal-Singer. R; Kanniess. F; Simpson. KJ; Gibson. A; Vessev. RSJ: Janicki. S: Magnussen. H; Jorres.
RA; Richter. K. (2005). Validation of the human ozone challenge model as a tool for assessing anti-
inflammatory drugs in early development. J Clin Pharmacol 45: 498-503.
6-278
-------
Hoppe. P: Peters. A; Rabe. G: Praml. G: Lindner. J; Jakobi. G: Fruhmann. G: Nowak. D. (2003).
Environmental ozone effects in different population subgroups. Int J Hyg Environ Health 206: 505-516.
http://dx.doi.org/10.1078/1438-4639-00250
Hoppe. P; Praml. G: Rabe. G: Lindner. J: Fruhmann. G: Kessel. R. (1995). Environmental ozone field study on
pulmonary and subjective responses of assumed risk groups. Environ Res 71: 109-121.
Horstman. DH; Ball. BA; Brown. J: Gerritv. T; Folinsbee. LJ. (1995). Comparison of pulmonary responses of
asthmatic and nonasthmatic subjects performing light exercise while exposed to a low level of ozone.
Toxicol Ind Health 11: 369-385.
Horstman. DH; Folinsbee. LJ: Ives. PJ: Abdul-Salaam. S: McDonnell. WF. (1990). Ozone concentration and
pulmonary response relationships for 6.6-hour exposures with five hours of moderate exercise to 0.08, 0.10,
and 0.12 ppm. Am J Respir Crit Care Med 142: 1158-1163.
Horvath. SM; Gliner. JA; Folinsbee. LJ. (1981). Adaptation to ozone: Duration of effect. Am Rev Respir Dis
123: 496-499.
Horvath. SM; Gliner. JA; Matsen-Twisdale. JA. (1979). Pulmonary function and maximum exercise responses
following acute ozone exposure. Aviat Space Environ Med 50: 901-905.
Hosseinpoor. AR; Forouzanfar. MH; Yunesian. M; Asghari. F; Naieni. KH; Farhood. D. (2005). Air pollution
and hospitalization due to angina pectoris in Tehran, Iran: A time-series study. Environ Res 99: 126-131.
Houslev. DG: Eccles. R; Richards. RJ. (1996). Gender difference in the concentration of the antioxidant uric
acid in human nasal lavage. Acta Otolaryngol 116: 751-754.
Howarth. PH; Persson. CG: Meltzer. EO; Jacobson. MR; Durham. SR; Silkoff. PE. (2005). Objective
monitoring of nasal airway inflammation in rhinitis [Review]. J Allergy Clin Immunol 115: S414-S441.
http://dx.doi.0rg/10.1016/i.jaci.2004.12.1134
Hunt. J. (2002). Exhaled breath condensate: An evolving tool for noninvasive evaluation of lung disease
[Review]. J Allergy Clin Immunol 110: 28-34. http://dx.doi.org/10.1067/mai.2002.124966
Hunt. JF; Fang. K; Malik. R; Snyder. A; Malhotra. N; Platts-Mills. TAB; Gaston. B. (2000). Endogenous
airway acidification: Implications for asthma pathophysiology. Am J Respir Crit Care Med 161: 694-699.
Hurst. DJ: Gardner. DE; Coffin. PL. (1970). Effect of ozone on acid hydrolases of the pulmonary alveolar
macrophage. J Reticuloendothel Soc 8: 288-300.
Hyde. DM; Hubbard. WC: Wong. V; Wu. R: Pinkerton. K; Plopper. CG. (1992). Ozone-induced acute
tracheobronchial epithelial injury: relationship to granulocyte emigration in the lung. Am J Respir Cell Mol
Biol 6: 481-497.
Inoue. K; Takano. H; Kaewamatawong. T; Shimada. A; Suzuki. J: Yanagisawa. R: Tasaka. S: Ishizaka. A;
Satoh. M. (2008). Role of metallothionein in lung inflammation induced by ozone exposure in mice. Free
Radic Biol Med 45: 1714-1722. http://dx.doi.Org/10.1016/i.freeradbiomed.2008.09.008
Ito. K; De Leon. SF; Lippmann. M. (2005). Associations between ozone and daily mortality, analysis and
meta-analysis. Epidemiology 16: 446-457.
Ito. K; Thurston. GD; Silverman. RA. (2007b). Characterization of PM2.5, gaseous pollutants, and
meteorological interactions in the context of time-series health effects models. J Expo Sci Environ
Epidemiol 17: S45-S60. http://dx.doi.org/10.1038/sj.jes.7500627
Iwasaki. T; Takahashi. M; Saito. H; Arito. H. (1998). Adaptation of extrapulmonary responses to ozone
exposure in conscious rats. Ind Health 36: 57-60.
Jakab. GJ: Hmieleski. RR. (1988). Reduction of influenza virus pathogenesis by exposure to 0.5 ppm ozone. J
Toxicol Environ Health 23: 455-472. http://dx.doi.org/10.1080/15287398809531128
Jalaludin. BB; Chev. T; O'Toole. BI; Smith. WT; Capon. AG: Leeder. SR. (2000). Acute effects of low levels
of ambient ozone on peak expiratory flow rate in a cohort of Australian children. Int J Epidemiol 29: 549-
557. http://dx.doi.Org/10.1093/iie/29.3.549
6-279
-------
Jalaludin. BB; O'Toole. BI; Leeder. S. R. (2004). Acute effects of urban ambient air pollution on respiratory
symptoms, asthma medication use, and doctor visits for asthma in a cohort of Australian children. Environ
Res 95: 32-42. http://dx.doi.org/10.1016/80013-9351(03)00038-0
Janero. DR. (1990). Malondialdehyde and thiobarbituric acid-reactivity as diagnostic indices of lipid
peroxidation and peroxidative tissue injury. Free Radic Biol Med 9: 515-540.
http://dx.doi.org/10.1016/0891-5849(90)90131-2
Jang. AS: Choi. IS: Yang. SY; Kim. YG: Lee. JH; Park. SW: Park. CS. (2005). Antioxidant responsiveness in
BALB/c mice exposed to ozone. Respiration 72: 79-84. http://dx.doi.org/10.1159/000083405
Johnston. RA; Mizgerd. JP; Flynt. L; Quinton. LJ; Williams. ES: Shore. SA. (2007). Type I interleukin-1
receptor is required for pulmonary responses to subacute ozone exposure in mice. Am J Respir Cell Mol
Biol 37: 477-484. http://dx.doi.org/10.1165/rcmb.2006-0315OC
Johnston. RA; Schwartzman. IN; Flynt. L; Shore. SA. (2005b). Role of interleukin-6 in murine airway
responses to ozone. Am J Physiol Lung Cell Mol Physiol 288: L390-L397.
http://dx.doi.org/10.1152/ajplung.00007.2004
Jones. SL; Kittelson. J: Cowan. JO; Flannery. EM; Hancox. RJ; McLachlan. CR; Taylor. DR. (2001).
The predictive value of exhaled nitric oxide measurements in assessing changes in asthma control. Am J
Respir Crit Care Med 164: 738-743.
Jorres. R; Nowak. D; Magnussen. H; Speckin. P; Koschyk. S. (1996). The effect of ozone exposure on allergen
responsiveness in subjects with asthma or rhinitis. Am J Respir Crit Care Med 153: 56-64.
Jorres. RA; Holz. O; Zachgo. W: Timm. P; Koschyk. S: Muller. B; Grimminger. F; Seeger. W: Kelly. FJ;
Punster. C: Frischer. T; Lubec. G: Waschewski. M; Niendorf. A; Magnussen. H. (2000). The effect of
repeated ozone exposures on inflammatory markers in bronchoalveolar lavage fluid and mucosal biopsies.
Am J Respir Crit Care Med 161: 1855-1861.
Just. J: Segala. C: Sahraoui. F; Priol. G: Grimfeld. A; Neukirch. F. (2002). Short-term health effects of
paniculate and photochemical air pollution in asthmatic children. Eur Respir J 20: 899-906.
http://dx.doi.org/10.1183/09031936.02.00236902
Katsouvanni. K; Samet. JM; Anderson. HR; Atkinson. R; Le Tertre. A; Medina. S: Samoli. E; Touloumi. G:
Burnett. RT; Krewski. D; Ramsay. T; Dominici. F; Peng. RD; Schwartz. J: Zanobetti. A. (2009). Air
pollution and health: A European and North American approach (APHENA). (Research Report 142).
Boston, MA: Health Effects Institute. http://pubs.healtheffects.org/view.php?id=327
Katsouvanni. K; Touloumi. G: Samoli. E; Grvparis. A; Le Tertre. A; Monopolis. Y; Rossi. G: Zmirou. D;
Ballester. F; Boumghar. A; Anderson. HR; Woityniak. B; Paldv. A; Braunstein. R; Pekkanen. J: Schindler.
C: Schwartz. J. (2001). Confounding and effect modification in the short-term effects of ambient particles
on total mortality: Results from 29 European cities within the APHEA2 project. Epidemiology 12: 521-531.
Kehrl. HR; Hazucha. MJ; Solic. JJ: Bromberg. PA. (1985). Responses of subjects with chronic obstructive
pulmonary disease after exposures to 0.3 ppm ozone. Am Rev Respir Dis 131: 719-724.
Kehrl. HR; Peden. DB; Ball. BA; Folinsbee. LJ: Horstman. DH. (1999). Increased specific airway reactivity of
persons with mild allergic asthma after 7.6 hours of exposure to 0.16 ppm ozone. J Allergy Clin Immunol
104: 1198-1204.
Kehrl. HR; Vincent. LM; Kowalsky. RJ: Horstman. DH; O'Neil. JJ: McCartney. WH; Bromberg. PA. (1987).
Ozone exposure increases respiratory epithelial permeability in humans. Am Rev Respir Dis 135: 1124-
1128.
Kharitonov. SA; Barnes. PJ. (2000). Clinical aspects of exhaled nitric oxide [Review]. Eur Respir J 16: 781-
792.
Khatri. SB; Holguin. FC: Ryan. PB; Manning. D; Erzurum. SC: Teague. WG. (2009). Association of ambient
ozone exposure with airway inflammation and allergy in adults with asthma. J Asthma 46: 777-785.
htrp://dx.doi.org/10.1080/02770900902779284
6-280
-------
Kim. CS: Alexis. NE; Rappold. AG: Kehrl. H; Hazucha. MJ; Lav. JC: Schmitt. MT; Case. M; Devlin. RB:
Peden. DB; Diaz-Sanchez. D. (2011). Lung function and inflammatory responses in healthy young adults
exposed to 0.06 ppm ozone for 6.6 hours. Am J Respir Crit Care Med 183: 1215-1221.
http://dx.doi.org/10.1164/rccm.201011-1813OC
Kinnev. PL; Thurston. GD; Raizenne. M. (1996). The effects of ambient ozone on lung function in children:
Areanalysis of six summer camp studies. Environ Health Perspect 104: 170-174.
Kleeberger. SR; Reddv. S: Zhang. LY; Jedlicka. AE. (2000). Genetic susceptibility to ozone-induced lung
hyperpermeability: Role of toll-like receptor 4. Am J Respir Cell Mol Biol 22: 620-627.
Klestadt. D; Laval-Gilly. P; Foucaud. L; Falla. J. (2005). Influences of ozone exposure upon macrophage
responsivity to N-formyl-methionyl-leucyl-phenylalanine: Mobility and metabolic changes. Toxicol
In Vitro 19: 199-206. http://dx.doi.Org/10.1016/i.tiv.2004.08.004
Kodavanti. UP; Thomas. R; Ledbetter. AD; Schladweiler. MC: Shannahan. JH; Wallenborn. JG: Lund. AK;
Campen. MJ: Butler. EO; Gottipolu. RR; Nyska. A; Richards. JE; Andrews. D: Jaskot. RH; Mckee. J:
Kotha. S. R.: Patel. RB; Parianandi. NL. (2011). Vascular and cardiac impairments in rats Inhaling ozone
and diesel exhaust particles. Environ Health Perspect 119: 312-318. http://dx.doi.org/10.1289/ehp.1002386
Koren. HS: Devlin. RB: Graham. DE; Mann. R; Mcgee. MP; Horstman. DH; Kozumbo. WJ: Becker. S:
House. DE; Mcdonnell. WF; Bromberg. PA. (1989). Ozone-induced inflammation in the lower airways of
human subjects. Am J Respir Crit Care Med 139: 407-415.
Korrick. SA; Neas. LM; Dockery. DW: Gold. PR; Allen. GA; Hill. LB; Kimball. KD; Rosner. BA; Speizer.
FE. (1998). Effects of ozone and other pollutants on the pulmonary function of adult hikers. Environ Health
Perspect 106: 93-99. http://dx.doi.org/10.1289/ehp.9810693
Kostikas. K; Papatheodorou. G: Ganas. K; Psathakis. K; Panagou. P; Loukides. S. (2002). pH in expired breath
condensate of patients with inflammatory airway diseases. Am J Respir Crit Care Med 165: 1364-1370.
Kreit. JW: Gross. KB; Moore. TB; Lorenzen. TJ; D'Arcv. J: Eschenbacher. WL. (1989). Ozone-induced
changes in pulmonary function and bronchial responsiveness in asthmatics. J Appl Physiol 66: 217-222.
Kulle. TJ; Sauder. LR; Hebel. JR; Chatham. MD. (1985). Ozone response relationships in healthy nonsmokers.
Am Rev Respir Dis 132: 36-41.
Kulle. TJ: Sauder. LR; Kerr. HP; Farrell. BP; Bermel. MS: Smith. DM. (1982). Duration of pulmonary
function adaptation to ozone in humans. Am Ind Hyg Assoc J 43: 832-837.
Lagorio. S: Forastiere. F; Pistelli. R: lavarone. I; Michelozzi. P; Fano. V; Marconi. A; Ziemacki. G: Ostro. BD.
(2006). Air pollution and lung function among susceptible adult subjects: A panel study. Environ Health 5:
11. http://dx.doi.org/10.1186/1476-069X-5-ll
Lanki. T; Pekkanen. J: Aalto. P: Elosua. R: Berglind. N; D'Ippoliti. D: Kulmala. M; Nyberg. F; Peters. A;
Picciotto. S: Salomaa. V: Sunyer. J: Tiittanen. P; Von Klot. S: Forastiere. F. (2006). Associations of traffic-
related air pollutants with hospitalisation for first acute myocardial infarction: The HEAPS S study. Occup
Environ Med 63: 844-851. http://dx.doi.org/10.1136/oem.2005.023911
Larrieu. S: Jusot. JF; Blanchard. M; Prouvost. H; Declercq. C: Fabre. P: Pascal. L; Le Tertre. A; Wagner. V;
Riviere. S: Chardon. B; Borelli. D: Cassadou. S: Eilstein. D; Lefranc. A. (2007). Short term effects of air
pollution on hospitalizations for cardiovascular diseases in eight French cities: The PSAS program. Sci
Total Environ 387: 105-112.
Larsen. ST; Matsubara. S: Mcconville. G: Poulsen. SS: Gelfand. EW. (2010). Ozone increases airway
hyperreactivity and mucus hyperproduction in mice previously exposed to allergen. J Toxicol Environ
Health A 73: 738-747. http://dx.doi.org/10.1080/15287391003614034
Laskin. PL; Heck. DE; Laskin. JD. (1998). Role of inflammatory cytokines and nitric oxide in hepatic and
pulmonary toxicity. Toxicol Lett 102-103: 289-293.
Laskin. PL; Laskin. JD. (2001). Role of macrophages and inflammatory mediators in chemically induced
toxicity [Review]. Toxicology 160: 111-118.
6-281
-------
Laskin. PL; Pendino. KJ; Punjabi. CJ: del Valle. MR; Laskin. JD. (1994). Pulmonary and hepatic effects of
inhaled ozone in rats. Environ Health Perspect 10: 61-64.
Laskin. JD; Heck. DE; Laskin. PL. (1996). Nitric oxide production in the lung and liver following inhalation
of the pulmonary irritant ozone. Adv Exp Med Biol 387: 141-146.
Last. JA; Gohil. K; Mathrani. VC: Kenyon. NJ. (2005). Systemic responses to inhaled ozone in mice: cachexia
and down-regulation of liver xenobiotic metabolizing genes. Toxicol Appl Pharmacol 208: 117-126.
http://dx.doi.0rg/10.1016/i.taap.2005.02.001
Last. JA; Reiser. KM; Tyler. WS: Rucker. RB. (1984). Long-term consequences of exposure to ozone. I. Lung
collagen content. Toxicol Appl Pharmacol 72: 111-118.
Lawrence. SO: Simpson-Haidaris. PJ. (2004). Regulated de novo biosynthesis of fibrinogen in extrahepatic
epithelial cells in response to inflammation. Thromb Haemostasis 92: 234-243.
http://dx.doi.org/10.1160/TH04-01-0024
Lav. JC: Alexis. NE; Kleeberger. SR; Roubev. RA; Harris. BD; Bromberg. PA; Hazucha. MJ; Devlin. RB;
Peden. DB. (2007). Ozone enhances markers of innate immunity and antigen presentation on airway
monocytes in healthy individuals [Letter]. J Allergy Clin Immunol 120: 719-722.
http://dx.doi.0rg/10.1016/i.jaci.2007.05.005
Lee. IM; Tsai. SS: Ho. CK; Chiu. HF; Wu. TN; Yang. CY. (2008a). Air pollution and hospital admissions for
congestive heart failure: Are there potentially sensitive groups? Environ Res 108: 348-353.
http://dx.doi.0rg/10.1016/i.envres.2008.07.024
Lee. IM; Tsai. SS: Ho. CK; Chiu. HF; Yang. CY. (2007). Air pollution and hospital admissions for congestive
heart failure in a tropical city: Kaohsiung, Taiwan. Inhal Toxicol 19: 899-904.
http://dx.doi.org/10.1080/08958370701479406
Lee. JT; Kim. H; Cho. YS: Hong. YC: Ha. EH; Park. H. (2003b). Air pollution and hospital admissions for
ischemic heart diseases among individuals 64+ years of age residing in Seoul, Korea. Arch Environ Health
58: 617-623.
Lefohn. AS: Hazucha. MJ: Shadwick. D; Adams. WC. (2010a). An alternative form and level of the human
health ozone standard. Inhal Toxicol 22: 999-1011. http://dx.doi.org/10.3109/08958378.2010.505253
Leonard. RJ: Charpied. GL; Faddis. B. (1991). Effects of ambient inhaled ozone on vocal fold mucosa in
Bonnet monkeys. J Voice 5: 304-309. http://dx.doi.org/10.1016/S0892-1997(05)80060-8
Lew. JI; Chemerynski. SM; Sarnat. JA. (2005). Ozone exposure and mortality, an empiric Bayes
metaregression analysis. Epidemiology 16: 458-468.
Lewis. TC: Robins. TG: Dvonch. JT: Keeler. GJ: Yip. FY: Mentz. GB: Lin. X: Parker. EA: Israel. BA:
Gonzalez. L; Hill. Y (2005). Air pollution-associated changes in lung function among asthmatic children in
Detroit. Environ Health Perspect 113: 1068-1075.
Liao. D: Duan. Y; Whitsel. EA; Zheng. ZJ; Heiss. G: Chinchilli. VM; Lin. HM. (2004a). Association of higher
levels of ambient criteria pollutants with impaired cardiac autonomic control: a population-based study. Am
JEpidemiol 159: 768-777.
Liao. D: Heiss. G: Chinchilli. VM; Duan. Y; Folsom. AR; Lin. HM; Salomaa. V. (2005). Association of criteria
pollutants with plasma hemostatic/inflammatory markers: A population-based study. J Expo Sci Environ
Epidemiol 15: 319-328.
Lim. Y; Phung. AD; Corbacho. AM; Aung. HH; Maioli. E; Reznick. AZ; Cross. CE; Davis. PA; Valacchi. G.
(2006). Modulation of cutaneous wound healing by ozone: Differences between young and aged mice.
Toxicol Lett 160: 127-134. http://dx.doi.Org/10.1016/j.toxlet.2005.06.013
Lin. S: Bell. EM; Liu. W: Walker. RJ: Kim. NK; Hwang. SA. (2008a). Ambient ozone concentration and
hospital admissions due to childhood respiratory diseases in New York State, 1991-2001. Environ Res 108:
42-47. http://dx.doi.0rg/10.1016/i.envres.2008.06.007
6-282
-------
Linares. C: Diaz. J. (2010). Short-term effect of concentrations of fine paniculate matter on hospital
admissions due to cardiovascular and respiratory causes among the over-75 age group in Madrid, Spain.
Public Health 124: 28-36. http://dx.doi.Org/10.1016/i.puhe.2009.ll.007
Linn. WS: Avol. EL; Shamoo. DA; Peng. RC: Valencia. LM; Little. DE; Hackney. JD. (1988). Repeated
laboratory ozone exposures of volunteer Los Angeles residents: an apparent seasonal variation in response.
Toxicol Ind Health 4: 505-520.
Linn. WS: Avol. EL; Shamoo. DA; Spier. CE; Valencia. LM; Venet TG: Fischer. DA; Hackney. JD. (1986).
A dose-response study of healthy, heavily exercising men exposed to ozone at concentrations near the
ambient air quality standard. Toxicol Ind Health 2: 99-112.
Linn. WS: Fischer. DA; Medwav. DA; Anzar. UT; Spier. CE; Valencia. LM: Venet. TG: Hackney. JD. (1982a).
Short-term respiratory effects of 0.12 ppm ozone exposure in volunteers with chronic obstructive
pulmonary disease. Am Rev Respir Dis 125: 658-663.
Linn. WS: Medwav. DA; Anzar. UT; Valencia. LM: Spier. CE: FS-D. T; Fischer. DA; Hackney. JD. (1982b).
Persistence of adaptation to ozone in volunteers exposed repeatedly for six weeks. Am Rev Respir Dis 125:
491-495.
Linn. WS: Shamoo. DA; Anderson. KR; Peng. RC: Avol. EL; Hackney. JD: Gong. H. Jr. (1996). Short-term air
pollution exposures and responses in Los Angeles area schoolchildren. J Expo Sci Environ Epidemiol 6:
449.472.
Linn. WS: Shamoo. DA; Venet. TG: Spier. CE: Valencia. LM: Anzar. UT: Hackney. JD. (1983). Response to
ozone in volunteers with chronic obstructive pulmonary disease. Arch Environ Occup Health 38: 278-283.
Linn. WS: Szlachcic. Y; Gong. H. Jr; Kinnev. PL; Berhane. KT. (2000). Air pollution and daily hospital
admissions in metropolitan Los Angeles. Environ Health Perspect 108: 427-434.
Lisabeth. LD; Escobar. JD: Dvonch. JT; Sanchez. BN; Maiersik. JJ; Brown. PL; Smith. MA; Morgenstern.
LB. (2008). Ambient air pollution and risk for ischemic stroke and transient ischemic attack. Ann Neurol
64: 53-59. http://dx.doi.org/10.1002/ana.21403
Liu. L; Leech. JA; Urch. RB; Poon. R; Zimmerman. B; Kubav. JM; Silverman. FS. (1999). A comparison of
biomarkers of ozone exposure in human plasma, nasal lavage, and sputum. Inhal Toxicol 11: 657-674.
http://dx.doi.org/10.1080/089583799196790
Liu. L; Leech. JA; Urch. RB; Silverman. FS. (1997). In vivo salicylate hyroxylation: A potential biomarker for
assessing acute ozone exposure and effects in humans. Am J Respir Crit Care Med 156: 1405-1412.
Liu. L; Poon. R; Chen. L; Frescura. AM; Montuschi. P; Ciabattoni. G: Wheeler. A; Dales. R. (2009a). Acute
effects of air pollution on pulmonary function, airway inflammation, and oxidative stress in asthmatic
children. Environ Health Perspect 117: 668-674. http://dx.doi.org/10.1289/ehpll813
Mann. JK; Balmes. JR: Bruckner. TA; Mortimer. KM; Margolis. HG: Pratt. B; Hammond. SK; Lurmann. F;
Tager. IB. (2010). Short-term effects of air pollution on wheeze in asthmatic children in Fresno, California.
Environ Health Perspect 118: 1497-1502. http://dx.doi.org/10.1289/ehp.0901292
Mapp. CE; Fryer. AA; De Marzo. N; Pozzato. V: Padoan. M; Boschetto. P; Strange. RC: Hemmingsen. A;
Spiteri. MA. (2002). Glutathione S-transferase GSTP1 is a susceptibility gene for occupational asthma
induced by isocyanates. J Allergy Clin Immunol 109: 867-872. http://dx.doi.org/10.1067/mai.2002.123234
Mar. TF; Koenig. JQ. (2009). Relationship between visits to emergency departments for asthma and ozone
exposure in greater Seattle, Washington. Ann Allergy Asthma Immunol 103: 474-479.
http://dx.doi.org/10.1016/S1081-1206(10)60263-3
Martinez-Canabal. A; Angora-Perez. M. (2008). Effect of growth hormone on cyclooxygenase-2 expression in
the hippocampus of rats chronically exposed to ozone. Int JNeurosci 118: 455-469.
http://dx.doi.org/10.1080/00207450701593160
Martrette. JM; Thornton. SN; Trabalon. M. (2011). Prolonged ozone exposure effects behaviour, hormones and
respiratory muscles in young female rats. Physiol Behav 103: 302-307.
http://dx.doi.0rg/10.1016/i.physbeh.2011.02.024
6-283
-------
McDonnell. WF. (1996). Individual variability in human lung function responses to ozone exposure. Environ
Toxicol Pharmacol 2: 171-175.
McDonnell. WF; Chapman. RS: Horstman. DH; Leigh. MW: Abdul-Salaam. S. (1985a). A comparison of the
responses of children and adults to acute ozone exposure.
McDonnell. WF: Horstman. DH; Hazucha. MJ; Seal. E. Jr; Haak. ED; Salaam. SA; House. DE. (1983).
Pulmonary effects of ozone exposure during exercise: Dose-response characteristics. J Appl Physiol 54:
1345-1352.
McDonnell. WF. Ill; Chapman. RS: Leigh. MW: Strope. GL; Collier. AM. (1985b). Respiratory responses of
vigorously exercising children to 0.12 ppm ozone exposure. Am Rev Respir Dis 132: 875-879.
McDonnell. WF. Ill; Horstman. DH: Abdul-Salaam. S: House. DE. (1985c). Reproducibility of individual
responses to ozone exposure. Am Rev Respir Dis 131: 36-40.
McDonnell. WF: Kehrl. HR; Abdul-Salaam. S: Ives. PJ: Folinsbee. LJ; Devlin. RB; OTSTeil. JJ: Horstman. DH.
(1991). Respiratory response of humans exposed to low levels of ozone for 6.6 hours. Arch Environ Occup
Health 46: 145-150.
McDonnell. WF: Stewart. PW: Andreoni. S: Seal. E. Jr; Kehrl. HR; Horstman. DH: Folinsbee. LJ: Smith. MV.
(1997). Prediction of ozone-induced FEV1 changes: Effects of concentration, duration, and ventilation. Am
J Respir Crit Care Med 156: 715-722.
Mcdonnell. WF: Stewart. PW: Smith. MV. (2007). The temporal dynamics of ozone-induced FEV1 changes in
humans: An exposure-response model. Inhal Toxicol 19: 483-494.
McDonnell. WF: Stewart. PW: Smith. MV. (2010). Prediction of ozone-induced lung function responses in
humans. Inhal Toxicol 22: 160-168. http://dx.doi.org/10.3109/08958370903089557
Mcdonnell. WF: Stewart. PW: Smith. MV: Kim. CS: Schelegle. ES. (2012). Prediction of lung function
response for populations exposed to a wide range of ozone conditions. Inhal Toxicol 24: 619-633.
http://dx.doi.org/10.3109/08958378.2012.705919
McDonnell. WF: Stewart. PW: Smith. MV: Pan. WK; Pan. J. (1999b). Ozone-induced respiratory symptoms:
Exposure-response models and association with lung function. Eur Respir J 14: 845-853.
Medina-Ramon. M; Schwartz. J. (2008). Who is more vulnerable to die from ozone air pollution?
Epidemiology 19: 672-679. http://dx.doi.org/10.1097/EDE.Ob013e3181773476
Medina-Ramon. M; Zanobetti. A; Schwartz. J. (2006). The effect of ozone and PM10 on hospital admissions
for pneumonia and chronic obstructive pulmonary disease: A national multicity study. Am J Epidemiol 163:
579-588. http://dx.doi.org/10.1093/aje/kwj078
Mellick. PW: Dungworth. PL; Schwartz. LW: Tyler. WS. (1977). Short term morphologic effects of high
ambient levels of ozone on lungs of rhesus monkeys. Lab Invest 36: 82-90.
Messineo. TD; Adams. WC. (1990). Ozone inhalation effects in females varying widely in lung size:
Comparison with males. J Appl Physiol 69: 96-103.
Metzger. KB; Klein. M; Flanders. WD; Peel. JL; Mulholland. JA; Langberg. JJ: Tolbert. PE. (2007). Ambient
air pollution and cardiac arrhythmias in patients with implantable defibrillators. Epidemiology 18: 585-592.
http://dx.doi.org/10.1097/EDE.Ob013e318124ffOe
Metzger. KB: Tolbert. PE: Klein. M: Peel. JL: Flanders. WD: Todd. KH: Mulholland. JA: Ryan. PB: Frumkin.
II (2004). Ambient air pollution and cardiovascular emergency department visits. Epidemiology 15: 46-56.
Middleton. N; Yiallouros. P; Kleanthous. S: Kolokotroni. O; Schwartz. J: Dockerv. DW: Demokritou. P;
Koutrakis. P. (2008). A 10-year time-series analysis of respiratory and cardiovascular morbidity in Nicosia,
Cyprus: The effect of short-term changes in air pollution and dust storms. Environ Health 7: 39.
Mikerov. AN; Umstead. TM; Gan. X; Huang. W: Quo. X; Wang. G: Phelps. PS: Floros. J. (2008c). Impact of
ozone exposure on the phagocytic activity of human surfactant protein A (SP-A) and SP-A variants. Am J
Physiol Lung Cell Mol Physiol 294: L121-L130. http://dx.doi.org/10.1152/ajplung.00288.2007
6-284
-------
Miller. FJ; Illing. JW: Gardner. DE. (1978). Effect of urban ozone levels on laboratory-induced respiratory
infections. Toxicol Lett 2: 163-169.
Mokoena. ML; Harvey. BH; Oliver. DW: Brink. CB. (2010). Ozone modulates the effects of imipramine on
immobility in the forced swim test, and nonspecific parameters of hippocampal oxidative stress in the rat.
Metab Brain Dis 25: 125-133. http://dx.doi.org/10.1007/sll011-010-9189-7
Molfino. NA; Wright. SC: Katz. I; Tarlo. S: Silverman. F; Mcclean. PA; Szalai. JP; Raizenne. M; Slutsky. AS:
Zamel. N. (1991). Effect of low concentrations of ozone on inhaled allergen responses in asthmatic
subjects. Lancet 338: 199-203.
Moon. JS: Kim. YS: Kim. JH; Son. BS: Kim. PS: Yang. W. (2009). Respiratory health effects among
schoolchildren and their relationship to air pollutants in Korea. Int J Environ Health Res 19: 31-48.
http://dx.doi.org/10.1080/09603120802272201
Morgan. G: Corbett. S: Wlodarczyk. J. (1998). Air pollution and hospital admissions in Sydney, Australia,
1990 to 1994. Am J Public Health 88: 1761-1766.
Morrison. D; Rahman. I; MacNee. W. (2006). Permeability, inflammation and oxidant status in airspace
epithelium exposed to ozone. Respir Med 100: 2227-2234. http://dx.doi.Org/10.1016/i.rmed.2005.10.005
Morrow. JD; Hill. KE; Burk. RF; Nammour. TM; Badr. KF; Roberts. LJ. 2nd. (1990). A series of prostaglandin
F2-like compounds are produced in vivo in humans by a non-cyclooxygenase, free radical-catalyzed
mechanism. PNAS 87: 9383-9387.
Mortimer. KM; Neas. LM; Dockery. DW: Redline. S: Tager. IB. (2002). The effect of air pollution on inner-
city children with asthma. Eur Respir J 19: 699-705. http://dx.doi.org/10.1183/09031936.02.00247102
Mortimer. KM; Tager. IB; Dockery. DW: Neas. LM; Redline. S. (2000). The effect of ozone on inner-city
children with asthma: Identification of susceptible subgroups. Am J Respir Crit Care Med 162: 1838-1845.
Mudwav. IS: Blomberg. A; Frew. AJ; Holgate. ST; Sandstrom. T; Kelly. FJ. (1999a). Antioxidant consumption
and repletion kinetics in nasal lavage fluid following exposure of healthy human volunteers to ozone. Eur
Respir J 13: 1429-1438.
Mudwav. IS: Kelly. FJ. (2004a). An investigation of inhaled ozone dose and the magnitude of airway
inflammation in healthy adults. Am J Respir Crit Care Med 169: 1089-1095.
Mudwav. IS: Stenfors. N; Blomberg. A; Helledav. R; Punster. C: Marklund. SL; Frew. AJ: Sandstrom. T;
Kelly. FJ. (2001). Differences in basal airway antioxidant concentrations are not predictive of individual
responsiveness to ozone: A comparison of healthy and mild asthmatic subjects. Free Radic Biol Med 31:
962-974.
Murphy. SD; Ulrich. CE; Frankowitz. SH; Xintaras. C. (1964). Altered function in animals inhaling low
concentrations of ozone and nitrogen dioxide. Am Ind Hyg Assoc J 25: 246-253.
Murugan. A; Prys-Picard. C: Calhoun. WJ. (2009). Biomarkers in asthma [Review]. Curr Opin Pulm Med 15:
12-18. http://dx.doi.org/10.1097/MCP.Ob013e32831de235
Naeher. LP; Holford. TR; Beckett. WS: Belanger. K; Triche. EW: Bracken. MB; Leaderer. BP (1999). Healthy
women's PEF variations with ambient summer concentrations of PM10, PM2.5, SO42-, H+, and O3. Am J
Respir Crit Care Med 160: 117-125.
Nakamura. K; Matsunaga. K. (1998). Susceptibility of natural killer (NK) cells to reactive oxygen species
(ROS) and their restoration by the mimics of superoxide dismutase (SOD). Cancer Biother Radiopharm 13:
275-290.
Neas. LM; Dockery. DW: Koutrakis. P; Speizer. FE. (1999). Fine particles and peak flow in children: Acidity
versus mass. Epidemiology 10: 550-553.
Neas. LM; Dockery. DW: Koutrakis. P; Tollerud. DJ: Speizer. FE. (1995). The association of ambient air
pollution with twice daily peak expiratory flow rate measurements in children. Am J Epidemiol 141: 111-
122.
6-285
-------
Neidell. M. (2009). Information, avoidance behavior, and health: The effect of ozone on asthma
hospitalizations. Journal of Human Resources 44: 450-478.
Neidell. M; Kinnev. PL. (2010). Estimates of the association between ozone and asthma hospitalizations that
account for behavioral responses to air quality information. Environ Sci Pol 13: 97-103.
http://dx.doi.0rg/10.1016/i.envsci.2009.12.006
Neuberger. M; Schimek. MG: Horak. F. Jr; Moshammer. H; Kundi. M; Frischer. T; Gomiscek. B; Puxbaum. H;
Hauck. H; AUPHEP-Team. (2004). Acute effects of paniculate matter on respiratory diseases, symptoms
and functions: Epidemiological results of the Austrian Projects on Health Effects of Paniculate Matter
(AUPHEP). Atmos Environ 38: 3971-3981.
Nickmilder. M; De Burbure. C: Sylviane. C: Xavier. D; Alfred. B; Alain. D. (2007). Increase of exhaled nitric
oxide in children exposed to low levels of ambient ozone. J Toxicol Environ Health A 70: 270-274.
Nightingale. JA; Rogers. DF; Chung. KF; Barnes. PJ. (2000). No effect of inhaled budesonide on the response
to inhaled ozone in normal subjects. Am J Respir Crit Care Med 161: 479-486.
O'Connor. GT; Neas. L; Vaughn. B; Kattan. M; Mitchell. H; Grain. EF; III. ER; Gruchalla. R; Morgan. W:
Stout. J: Adams. GK; Lippmann. M. (2008). Acute respiratory health effects of air pollution on children
with asthma in US inner cities. J Allergy Clin Immunol 121: 1133-1139.
http://dx.doi.0rg/10.1016/i.iaci.2008.02.020
O'Neill. MS: Ramirez-Aguilar. M; Meneses-Gonzalez. F; Hernandez-Avila. M; Geyh. AS: Sienra-Monge. JJ:
Romieu. I. (2003). Ozone exposure among Mexico City outdoor workers. J Air Waste Manag Assoc 53:
339-346.
Orazzo. F; Nespoli. L; Ito. K; Tassinari. D; Giardina. D: Funis. M; Cecchi. A; Trapani. C: Forgeschi. G:
Vignini. M; Nosetti. L; Pigna. S: Zanobetti. A. (2009). Air pollution, aeroallergens, and emergency room
visits for acute respiratory diseases and gastroenteric disorders among young children in six Italian cities.
Environ Health Perspect 117: 1780-1785. http://dx.doi.org/10.1289/ehp.0900599
Ostro. B; Lipsett. M; Mann. J: Braxton-Owens. H; White. M. (2001). Air pollution and exacerbation of asthma
in African-American children in Los Angeles. Epidemiology 12: 200-208.
Oudin. A; Stromberg. U; Jakobsson. K; Stroh. E; Bjork. J. (2010). Estimation of short-term effects of air
pollution on stroke hospital admissions in southern Sweden. Neuroepidemiology 34: 131-142.
http://dx.doi.org/10.1159/000274807
Ovarzun. M; Dussaubat N; Gonzalez. S. (2005). Effect of 0.25 ppm ozone exposure on pulmonary damage
induced by bleomycin. Biol Res 38: 353-358.
Park. JW: Lim. YH; Kyung. SY; An. CH; Lee. SP; Jeong. SH; Ju. SY. (2005a). Effects of ambient particulate
matter on peak expiratory flow rates and respiratory symptoms of asthmatics during Asian dust periods in
Korea. Respirology 10: 470-476. http://dx.doi.Org/10.llll/i.1440-1843.2005.00728.x
Park. SK; O'Neill. MS: Stunder. BJB; Vokonas. PS: Sparrow. D: Koutrakis. P; Schwartz. J. (2007). Source
location of air pollution and cardiac autonomic function: Trajectory cluster analysis for exposure
assessment. J Expo Sci Environ Epidemiol 17: 488-497. http://dx.doi.org/10.1038/sj.jes.7500552
Park. SK; O'Neill. MS: Vokonas. PS: Sparrow. D: Schwartz. J. (2005b). Effects of air pollution on heart rate
variability: The VANormative Aging Study. Environ Health Perspect 113: 304-309.
Park. SK: O'Neill. MS: Vokonas. PS: Sparrow. D: Wright. RO: Coull. B: Nie. H: Hu. H: Schwartz. J. (2008).
Air pollution and heart rate variability: Effect modification by chronic lead exposure. Epidemiology 19:
111-120. http://dx.doi.org/10.1097/EDE.Ob013e31815c408a
Passannante. AN; Hazucha. MJ; Bromberg. PA; Seal. E; Folinsbee. L; Koch. G. (1998). Nociceptive
mechanisms modulate ozone-induced human lung function decrements. J Appl Physiol 85: 1863-1870.
Peacock. JL; Anderson. HR; Bremner. SA; Marston. L; Seemungal. TA; Strachan. DP; Wedzicha. JA. (2011).
Outdoor air pollution and respiratory health in patients with COPD. Thorax 66: 591-596.
http://dx.doi.org/10.1136/tfax.2010.155358
6-286
-------
Peden. DB. (2011). The role of oxidative stress and innate immunity in O3 and endotoxin-induced human
allergic airway disease [Review]. Immunol Rev 242: 91-105. http://dx.doi.org/10.1111/i.1600-
065X.2011.01035.x
Peden. DB; Boehlecke. B; Horstman. D; Devlin. R. (1997). Prolonged acute exposure to 0.16 ppm ozone
induces eosinophilic airway inflammation in asthmatic subjects with allergies. J Allergy Clin Immunol 100:
802-808.
Peel. JL; Metzger. KB; Klein. M; Flanders. WD; Mulholland. JA; Tolbert. PE. (2007). Ambient air pollution
and cardiovascular emergency department visits in potentially sensitive groups. Am J Epidemiol 165: 625-
633. http://dx.doi.org/10.1093/aie/kwk051
Pellegrino. R; Viegi. G: Brusasco. V: Crapo. RO: Burgos. F; Casaburi. R; Coates. A; van der Grinten. CP;
Gustafsson. P; Hankinson. J: Jensen. R; Johnson. DC: Maclntyre. N; McKay. R; Miller. MR; Navajas. D;
Pedersen. OF; Wanger. J. (2005). Interpretative strategies for lung function tests. Eur Respir J 26: 948-968.
http://dx.doi.org/10.1183/09031936.05.00035205
Peng. RD; Dominici. F; Pastor-Barriuso. R: Zeger. SL; Samet. JM. (2005). Seasonal analyses of air pollution
and mortality in 100 US cities. 161: 585-594. http://dx.doi.org/10.1093/aje/kwi075
Perepu. RS: Garcia. C: Postal. D: Sethi. R. (2010). Enhanced death signaling in ozone-exposed ischemic-
reperfused hearts. Mol Cell Biochem 336: 55-64. http://dx.doi.org/10.1007/sll010-009-0265-4
Pereyra-Munoz. N; Rugerio-Vargas. C: Angoa-Perez. M; Borgonio-Perez. G: Rivas-Arancibia. S. (2006).
Oxidative damage in substantia nigra and striatum of rats chronically exposed to ozone. J Chem Neuroanat
31: 114-123. http://dx.doi.0rg/10.1016/i.jchemneu.2005.09.006
Peters. A; Dockery. DW: Muller. JE; Mittleman. MA. (2001). Increased paniculate air pollution and the
triggering of myocardial infarction. Circulation 103: 2810-2815.
http://dx.doi.org/10.1161/01.CIR.103.23.2810
Peterson. DC: Andrews. HL. (1963). The role of ozone in radiation avoidance in the mouse. Radiat Res 19:
331-336.
Petroeschevsky. A; Simpson. RW: Thalib. L; Rutherford. S. (2001). Associations between outdoor air pollution
and hospital admissions in Brisbane, Australia. Arch Environ Occup Health 56: 37-52.
Plopper. CG: Mango. GW: Hatch. GE; Wong. VJ; Toskala. E; Reynolds. SD; Tarkington. BK; Stripp. BR.
(2006). Elevation of susceptibility to ozone-induced acute tracheobronchial injury in transgenic mice
deficient in Clara cell secretory protein. Toxicol Appl Pharmacol 213: 74-85.
http://dx.doi.0rg/10.1016/i.taap.2005.09.003
Poloniecki. JD; Atkinson. RW: Ponce de Leon. A; Anderson. HR. (1997). Daily time series for cardiovascular
hospital admissions and previous day's air pollution in London, UK. Occup Environ Med 54: 535-540.
Power. K; Balmes. J: Solomon. C. (2008). Controlled exposure to combined particles and ozone decreases
heart rate variability. J Occup Environ Med 50: 1253-1260.
http://dx.doi.org/10.1097/JOM.Ob013e3181814239
Prescott. GJ: Cohen. GR; Elton. RA; Fowkes. FOR; Agius. RM. (1998). Urban air pollution and
cardiopulmonary ill health: A 145 year time series study. Occup Environ Med 55: 697-704.
Pulfer. MK; Murphy. RC. (2004). Formation of biologically active oxysterols during ozonolysis of cholesterol
present in lung surfactant. J Biol Chem 279: 26331-26338. http://dx.doi.org/10.1074/ibc.M403581200
Pulfer. MK; Taube. C: Gelfand. E; Murphy. RC. (2005). Ozone exposure in vivo and formation of biologically
active oxysterols in the lung. J Pharmacol Exp Ther 312: 256-264.
http://dx.doi.org/10.1124/ipet.104.073437
Qian. Z; Lin. HM; Chinchilli. VM; Lehman. EB; Duan. Y; Craig. TJ; Wilson. WE; Liao. D: Lazarus. SC:
Bascom. R. (2009). Interaction of ambient air pollution with asthma medication on exhaled nitric oxide
among asthmatics. Arch Environ Occup Health 64: 168-176.
http://dx.doi.org/10.1080/19338240903240616
6-287
-------
Que. LG: Stiles. JV: Sundv. JS: Foster. WM. (2011). Pulmonary function, bronchial reactivity, and epithelial
permeability are response phenotypes to ozone and develop differentially in healthy humans. J Appl
Physiol 111: 679-687. http://dx.doi.org/10.1152/iapplphysiol.00337.2011
Rabinovitch. N; Zhang. LN; Murphy. JR; Vedal. S: Dutton. SJ: Gelfand. EW. (2004). Effects of wintertime
ambient air pollutants on asthma exacerbations in urban minority children with moderate to severe disease.
J Allergy Clin Immunol 114: 1131-1137. http://dx.doi.Org/10.1016/i.iaci.2004.08.026
Raizenne. ME; Burnett. RT; Stern. B; Franklin. CA; Spengler. JD. (1989). Acute lung function responses to
ambient acid aerosol exposures in children. Environ Health Perspect 79: 179-185.
Ramirez-Aguilar. M; Barraza-Villarreal. A; Moreno-Macias. H; Winer. AM; Cicero-Fernandez. P; Velez-
Marquez. MG: Cortez-Lugo. M; Sienra-Monge. JJ: Romieu. I. (2008). Assessment of personal exposure to
ozone in asthmatic children residing in Mexico City. Salud Publica Mex 50: 67-75.
Ren. C: Williams. GM; Mengersen. K; Morawska. L; Tong. S. (2008). Does temperature modify short-term
effects of ozone on total mortality in 60 large eastern US communities? An assessment using the NMMAPS
data. Environ Int 34: 451-458.
Revis. NW: Major. T; Dalbev. WE. (1981). Cardiovascular effects of ozone and cadmium inhalation in the rat.
In Proceedings of the research planning workshop on health effects of oxidants (pp. 171-179). (EPA-600/9-
81-001). Research Triangle Park, NC: U.S. Environmental Protection Agency.
Rich. DQ; Kim. MH; Turner. JR; Mittleman. MA; Schwartz. J: Catalano. PJ; Dockery. DW. (2006a).
Association of ventricular arrhythmias detected by implantable cardioverter defibrillator and ambient air
pollutants in the St Louis, Missouri metropolitan area. Occup Environ Med 63: 591-596.
http://dx.doi.org/10.1136/oem.2005.023457
Puch. DQ; Kipen. HM; Zhang. J: Kamat. L; Wilson. AC: Kostis. JB. (2010). Triggering of transmural
infarctions, but not nontransmural infarctions, by ambient fine particles. Environ Health Perspect 118:
1229-1234. http://dx.doi.org/10.1289/ehp.0901624
Puch. DQ: Mittleman. MA; Link. MS: Schwartz. J: Luttmann-Gibson. H; Catalano. PJ: Speizer. FE; Gold. PR;
Dockery. DW. (2006b). Increased risk of paroxysmal atrial fibrillation episodes associated with acute
increases in ambient air pollution. Environ Health Perspect 114: 120-123.
http://dx.doi.org/10.1289/ehp.8371
Rich. DQ: Schwartz. J: Mittleman. MA; Link. M; Luttmann-Gibson. H; Catalano. PJ: Speizer. FE; Dockery.
DW. (2005). Association of short-term ambient air pollution concentrations and ventricular arrhythmias.
Am JEpidemiol 161: 1123-1132.
Riediker. M; Monn. C: Koller. T; Stahel. WA; Wuthrich. B. (2001). Air pollutants enhance rhinoconjunctivitis
symptoms in pollen-allergic individuals. Ann Allergy Asthma Immunol 87: 311-318.
http://dx.doi.org/10.1016/S1081-1206(10)62246-6
Rivas-Arancibia. S: Dorado-Martinez. C: Borgonio-Perez. G: Hiriart-Urdanivia. M; Verdugo-Diaz. L; Duran-
Vazquez. A; Colin-Baranque. L; Avila-Costa. MR. (2000). Effects of taurine on ozone-induced memory
deficits and lipid peroxidation levels in brains of young, mature, and old rats. Environ Res 82: 7-17.
http://dx.doi.org/10.1006/enrs.1999.3996
Rivas-Arancibia. S: Guevara-Guzman. R; Lopez-Vidal. Y; Rodriguez-Martinez. E; Gomes. MZ; Angoa-Perez.
M; Raisman-Vozari. R. (2010). Oxidative stress caused by ozone exposure induces loss of brain repair in
the hippocampus of adult rats. Toxicol Sci 113: 187-197. http://dx.doi.org/10.1093/toxsci/kfp252
Rivas-Arancibia. S: Vazquez-Sandoval. R: Gonzalez-Kladiano. D; Schneider-Rivas. S: Lechuga-Guerrero. A.
(1998). Effects of ozone exposure in rats on memory and levels of brain and pulmonary superoxide
dismutase. Environ Res 76: 33-39.
Rodriguez. C: Tonkin. R; Hevworth. J: Kusel. M; De Klerk. N; Sly. PD; Franklin. P; Runnion. T; Blocklev. A;
Landau. L; Hinwood. AL. (2007). The relationship between outdoor air quality and respiratory symptoms
in young children. Int J Environ Health Res 17: 351-360. http://dx.doi.org/10.1080/09603120701628669
6-288
-------
Romieu. I; Barraza-Villarreal. A; Escamilla-Nunez. C: Almstrand. AC: Diaz-Sanchez. D; Sly. PD; Olin. AC.
(2008). Exhaled breath malondialdehyde as a marker of effect of exposure to air pollution in children with
asthma. J Allergy Clin Immunol 121: 903-909. http://dx.doi.Org/10.1016/i.jaci.2007.12.004
Romieu. I; Barraza-Villarreal. A; Escamilla-Nunez. C: Texcalac-Sangrador. JL; Hernandez-Cadena. L; Diaz-
Sanchez. D; De Batlle. J: Del Rio-Navarro. BE. (2009). Dietary intake, lung function and airway
inflammation in Mexico City school children exposed to air pollutants. Respir Res 10: 122.
Romieu. I; Meneses. F; Ramirez. M; Ruiz. S: Padilla. RP; Sienra. JJ: Gerber. M; Grievink. L; Dekker. R;
Walda. I; Brunekreef. B. (1998b). Antioxidant supplementation and respiratory functions among workers
exposed to high levels of ozone. Am J Respir Crit Care Med 158: 226-232.
Romieu. I; Meneses. F; Ruiz. S: Huerta. J: Sienra. JJ: White. M; Etzel. R; Hernandez. M. (1997). Effects of
intermittent ozone exposure on peak expiratory flow and respiratory symptoms among asthmatic children
in Mexico City. Arch Environ Occup Health 52: 368-376.
Romieu. I; Meneses. F; Ruiz. S: Sienra. JJ: Huerta. J: White. MC: Etzel. RA. (1996). Effects of air pollution on
the respiratory health of asthmatic children living in Mexico City. Am J Respir Crit Care Med 154: 300-
307.
Romieu. I; Ramirez-Aguilar. M; Sienra-Monge. JJ: Moreno-Macias. H; Del Rio-Navarro. BE; David. G:
Marzec. J: Hernandez-Avila. M; London. S. (2006). GSTM1 and GSTP1 and respiratory health in
asthmatic children exposed to ozone. Eur Respir J 28: 953-959.
http://dx.doi.org/10.1183/09031936.06.00114905
Romieu. I; Sienra-Monge. JJ: Ramirez-Aguilar. M; Moreno-Macias. H; Reyes-Ruiz. NI; Estela del Rio-
Navarro. B; Hernandez-Avila. M; London. SJ. (2004b). Genetic polymorphism of GSTM1 and antioxidant
supplementation influence lung function in relation to ozone exposure in asthmatic children in Mexico City.
Thorax 59: 8-10.
Romieu. I; Sienra-Monge. JJ: Ramirez-Aguilar. M; Tellez-Rojo. MM; Moreno-Macias. H; Reyes-Ruiz. NI; Del
Rio-Navarro. BE; Ruiz-Navarro. MX; Hatch. G: Slade. R: Hernandez-Avila. M. (2002). Antioxidant
supplementation and lung functions among children with asthma exposed to high levels of air pollutants.
Am J Respir Crit Care Med 166: 703-709.
Rosenfeld. MA; Leonova. VB; Konstantinova. ML; Razumovskii. SD. (2009). Self-assembly of fibrin
monomers and fibrinogen aggregation during ozone oxidation. Biochemistry (Mosc) 74: 41-46.
http://dx.doi.org/10.1134/S0006297909010064
Ross. MA; Persky. VW: Scheff. PA; Chung. J: Curtis. L; Ramakrishnan. V; Wadden. RA; Hryhorczuk. DO.
(2002). Effect of ozone and aeroallergens on the respiratory health of asthmatics. Arch Environ Occup
Health 57: 568-578. http://dx.doi.org/10.1080/00039890209602090
Rozenfeld. MA; Leonova. VB; Konstantinova. ML; Razumovskii. SD: Makarov. VA; Nevedrova. OE;
Belozerskaia. GG. (2008). Disturbance of functional properties of fibrinogen under ozone oxidation. Dokl
Biochem Biophys 422: 315-318. http://dx.doi.org/10.1134/S1607672908050165
Rudez. G: Janssen. NA; Kilinc. E; Leebeek. FW: Gerlofs-Nijland. ME; Spronk. HM; ten Gate. H; Cassee. FR;
de Maat. MR (2009). Effects of ambient air pollution on hemostasis and inflammation. Environ Health
Perspect 117: 995-1001. http://dx.doi.org/10.1289/ehp.0800437
Ruidavets. JB; Cassadou. S: Cournot. M; Bataille. V: Mevbeck. M; Ferrieres. J. (2005a). Increased resting
heart rate with pollutants in a population based study. J Epidemiol Community Health 59: 685-693.
Ruidavets. JB; Cournot. M; Cassadou. S: Giroux. M; Mevbeck. M; Ferrieres. J. (2005b). Ozone air pollution is
associated with acute myocardial infarction. Circulation 111: 563-569.
Samet. JM; Hatch. GE; Horstman. D; Steck-Scott. S: Arab. L; Bromberg. PA; Levine. M; McDonnell. WF;
Devlin. RB. (2001). Effect of antioxidant supplementation on ozone-induced lung injury in human subjects.
Am J Respir Crit Care Med 164: 819-825.
6-289
-------
Samet JM; Zeger. SL; Dominici. F; Curriero. F; Coursac. I; Dockerv. DW: Schwartz. J; Zanobetti. A. (2000).
The national morbidity, mortality, and air pollution study. Part II: Morbidity, mortality, and air pollution in
the United States. Cambridge, MA: Health Effects Institute.
Samoli. E; Zanobetti. A; Schwartz. J; Atkinson. R; Le Tertre. A; Schindler. C: Perez. L; Cadum. E; Pekkanen.
J; Paldv. A; Touloumi. G: Katsouvanni. K. (2009). The temporal pattern of mortality responses to ambient
ozone in the APHEA project. J Epidemiol Community Health 63: 960-966.
http://dx.doi.org/10.1136/iech.2008.084012
Santiago-Lopez. D; Bautista-Martinez. JA; Reyes-Hernandez. CI; Aguilar-Martinez. M; Rivas-Arancibia. S.
(2010). Oxidative stress, progressive damage in the substantia nigra and plasma dopamine oxidation, in rats
chronically exposed to ozone. Toxicol Lett 197: 193-200. http://dx.doi.Org/10.1016/i.toxlet.2010.05.020
Santucci. D; Sorace. A; Francia. N; Aloe. L; Alleva. E. (2006). Prolonged prenatal exposure to low-level ozone
affects aggressive behaviour as well as NGF and BDNF levels in the central nervous system of CD-I mice.
Behav Brain Res 166: 124-130. http://dx.doi.Org/10.1016/i.bbr.2005.07.032
Sarnat. SB; Coull. BA; Schwartz. J: Gold. PR; Suh. HH. (2006a). Factors affecting the association between
ambient concentrations and personal exposures to particles and gases. Environ Health Perspect 114: 649-
654.
Sarnat. SE; Suh. HH; Coull. BA; Schwartz. J: Stone. PH; Gold. DR. (2006b). Ambient paniculate air pollution
and cardiac arrhythmia in a panel of older adults in Steubenville, Ohio. Occup Environ Med 63: 700-706.
http://dx.doi.org/10.1136/oem.2006.027292
Sathishkumar. K; Gao. X; Raghavamenon. AC: Parinandi. N; Pryor. WA; Uppu. RM. (2009). Cholesterol
secoaldehyde induces apoptosis in H9c2 cardiomyoblasts through reactive oxygen species involving
mitochondria! and death receptor pathways. Free Radic Biol Med 47: 548-558.
http://dx.doi.0rg/10.1016/i.freeradbiomed.2009.05.020
Sathishkumar. K; Hague. M; Perumal. TE; Francis. J: Uppu. RM. (2005). A major ozonation product of
cholesterol, 3beta-hydroxy-5-oxo-5,6-secocholestan-6-al, induces apoptosis in H9c2 cardiomyoblasts.
FEES Lett 579: 6444-6450. http://dx.doi.Org/10.1016/i.febslet.2005.10.044
Scannell. C: Chen. L; Aris. RM; Tager. I; Christian. D; Ferrando. R; Welch. B; Kelly. T; Balmes. JR. (1996).
Greater ozone-induced inflammatory responses in subjects with asthma. Am J Respir Crit Care Med 154:
24-29.
Scarlett. JF; Abbott. KJ; Peacock. JL; Strachan. DP; Anderson. HR. (1996). Acute effects of summer air
pollution on respiratory function in primary school children in southern England. Thorax 51: 1109-1114.
Schelegle. ES: Adams. WC. (1986). Reduced exercise time in competitive simulations consequent to low level
ozone exposure. Med Sci Sports Exerc 18: 408-414.
Schelegle. ES: Adams. WC: Walbv. WF; Marion. MS. (2012). Modelling of individual subject ozone exposure
response kinetics. Inhal Toxicol 24: 401-415. http://dx.doi.org/10.3109/08958378.2012.683891
Schelegle. ES: Morales. CA; Walbv. WF; Marion. S: Allen. RP. (2009). 6.6-hour inhalation of ozone
concentrations from 60 to 87 parts per billion in healthy humans. Am J Respir Crit Care Med 180: 265-272.
http://dx.doi.org/10.1164/rccm.200809-1484OC
Schelegle. ES: Siefkin. AD; McDonald. RJ. (1991). Time course of ozone-induced neutrophilia in normal
humans. Am J Respir Crit Care Med 143: 1353-1358.
Schildcrout. JS: Sheppard. L; Lumlev. T; Slaughter. JC: Koenig. JO; Shapiro. GG. (2006). Ambient air
pollution and asthma exacerbations in children: An eight-city analysis. Am J Epidemiol 164: 505-517.
http://dx.doi.org/10.1093/aie/kwi225
Schmekel. B; Ahlner. J: Malmstrom. M; Venge. P. (2001). Eosinophil cationic protein (ECP) in saliva: Anew
marker of disease activity in bronchial asthma. Respir Med 98: 670-675.
http://dx.doi.org/10.1053/rmed.2001.1123
Schwartz. J. (2005a). How sensitive is the association between ozone and daily deaths to control for
temperature? Am J Respir Crit Care Med 171: 627-631. http://dx.doi.org/10.1164/rccm.200407-933OC
6-290
-------
Schwartz. J. (2005b). Who is sensitive to extremes of temperature? A case-only analysis. Epidemiology 16: 67-
72. http://dx.doi.org/10.1097/01.ede.0000147114.25957.71
Schwartz. J; Litonjua. A; Suh. H; Verrier. M; Zanobetti. A; Syring. M; Nearing. B; Verrier. R; Stone. P;
Maccallum. G: Speizer. FE; Gold. DR. (2005). Traffic related pollution and heart rate variability in a panel
of elderly subjects. Thorax 60: 455-461. http://dx.doi.org/10.1136/thx.2004.024836
Schwartz. LW. (1976). Comparison of the effects of ozone and oxygen on lungs of rats. Environ Health
Perspect 16: 179-180.
Schwartz. LW: Dungworth. PL; Mustafa. MG: Tarkington. BK; Tyler. WS. (1976). Pulmonary responses of
rats to ambient levels of ozone: effects of 7-day intermittent or continuous exposure. Lab Invest 34: 565-
578.
Seal. E. Jr; McDonnell. WF; House. DE. (1996). Effects of age, socioeconomic status, and menstrual cycle on
pulmonary response to ozone. Arch Environ Occup Health 51: 132-137.
Seal. E. Jr; McDonnell. WF; House. DE; Salaam. SA; Dewitt PJ: Butler. SO: Green. J: Raggio. L. (1993).
The pulmonary response of white and black adults to six concentrations of ozone. Am J Respir Crit Care
Med 147: 804-810.
Selgrade. MK; Daniels. MJ; Grose. EC. (1990). Acute, subchronic, and chronic exposure to a simulated urban
profile of ozone: Effects on extrapulmonary natural killer cell activity and lymphocyte mitogenic responses.
InhalToxicol 2: 375-389.
Selwvn. BJ; Stock. TH; Hardy. RJ: Chan. FA; Jenkins. DE; Kotchmar. DJ; Chapman. RS. (1985). Health
effects of ambient ozone exposure in vigorously exercising adults. In Evaluation of the scientific basis for
ozone/oxidants standards: Proceedings of an apca international specialty conference. Pittsburgh, PA: Air
Pollution Control Association.
Sharkhuu. T; Doerfler. PL; Copeland. C: Luebke. RW: Gilmour. MI. (2011). Effect of maternal exposure to
ozone on reproductive outcome and immune, inflammatory, and allergic responses in the offspring. J
Immunotoxicol 8: 183-194. http://dx.doi.org/10.3109/1547691X.2011.568978
Sienra-Monge. JJ: Ramirez-Aguilar. M; Moreno-Macias. H; Reyes-Ruiz. NI; Del Rio-Navarro. BE; Ruiz-
Navarro. MX; Hatch. G: Crissman. K; Slade. R; Devlin. RB; Romieu. I. (2004). Antioxidant
supplementation and nasal inflammatory responses among young asthmatics exposed to high levels of
ozone. Clin Exp Immunol 138: 317-322. http://dx.doi.Org/10.llll/i.1365-2249.2004.02606.x
Silverman. RA; Ito. K. (2010). Age-related association of fine particles and ozone with severe acute asthma in
New York City. J Allergy Clin Immunol 125: 367-373. http://dx.doi.Org/10.1016/i.jaci.2009.10.061
Silverman. RA; Ito. K; Freese. J: Kaufman. BJ: De Claro. D; Braun. J: Prezant. DJ. (2010). Association of
ambient fine particles with out-of-hospital cardiac arrests in New York City. Am J Epidemiol 172: 917-923.
http://dx.doi.org/10.1093/aie/kwq217
Simonian. NA; Coyle. JT. (1996). Oxidative stress in neurodegenerative diseases [Review]. Annu Rev
Pharmacol Toxicol 36: 83-106. http://dx.doi.org/10.1146/annurev.pa.36.040196.000503
Simpson. R; Williams. G: Petroeschevsky. A; Best. T; Morgan. G: Denison. L; Hinwood. A; Neville. G.
(2005). The short-term effects of air pollution on hospital admissions in four Australian cities. Aust N Z J
Public Health 29: 213-221.
Sinclair. AH; Edgerton. ES: Wyzga. R: Tolsma. D. (2010). A two-time-period comparison of the effects of
ambient air pollution on outpatient visits for acute respiratory illnesses. J Air Waste Manag Assoc 60: 163-
175. http://dx.doi.0rg/10.3155/1047-3289.60.2.163
Sinclair. AH; Tolsma. D. (2004). Associations and lags between air pollution and acute respiratory visits in an
ambulatory care setting: 25-month results from the aerosol research and inhalation epidemiological study. J
Air Waste Manag Assoc 54: 1212-1218.
6-291
-------
Sivagangabalan. G: Spears. D; Masse. S: Urch. B; Brook. RD; Silverman. F; Gold. PR; Lukic. KZ; Speck. M;
Kusha. M; Farid. T; Poku. K; Shi. E; Floras. J: Nantfaakumar. K. (2011). The effect of air pollution on
spatial dispersion of myocardial repolarization in healthy human volunteers. J Am Coll Cardiol 57: 198-
206. http://dx.doi.0rg/10.1016/j.jacc.2010.08.625
Smith. RL; Xu. B; Switzer. P. (2009b). Reassessing the relationship between ozone and short-term mortality in
U.S. urban communities. Inhal Toxicol 21: 37-61. http://dx.doi.org/10.1080/08958370903161612
Solic. JJ: Hazucha. MJ; Bromberg. PA. (1982). The acute effects of 0.2 ppm ozone in patients with chronic
obstructive pulmonary disease. Am Rev Respir Dis 125: 664-669.
Son. JY; Bell. ML; Lee. JT. (2010). Individual exposure to air pollution and lung function in Korea: Spatial
analysis using multiple exposure approaches. Environ Res 110: 739-749.
http://dx.doi.0rg/10.1016/i.envres.2010.08.003
Soulage. C: Perrin. D; Cottet-Emard. JM; Pequignot. J: Dalmaz. Y; Pequignot. JM. (2004). Central and
peripheral changes in catecholamine biosynthesis and turnover in rats after a short period of ozone
exposure. Neurochem Int 45: 979-986. http://dx.doi.Org/10.1016/i.neuint.2004.06.015
Spannhake. EW: Reddv. SPM; Jacobv. DB; Yu. XY; Saatian. B; Tian. J. (2002). Synergism between rhinovirus
infection and oxidant pollutant exposure enhances airway epithelial cell cytokine production. Environ
Health Perspect 110: 665-670.
Spektor. DM; Lippmann. M. (1991). Health effects of ambient ozone on healthy children at a summer camp.
In RL Berglund; DR Lawson; DJ McKee (Eds.), Tropospheric Ozone and the Environment: Papers from an
International Conference; March 1990; Los Angeles, CA (pp. 83-89). Pittsburgh, PA: Air & Waste
Management Association.
Spektor. DM; Lippmann. M; Lioy. PJ; Thurston. GD; Citak. K; James. DJ; Bock. N; Speizer. FE; Haves. C.
(1988a). Effects of ambient ozone on respiratory function in active, normal children. Am Rev Respir Dis
137:313-320.
Spektor. DM; Lippmann. M; Thurston. GD; Lioy. PJ; Stecko. J; O'Connor. G; Garshick. E; Speizer. FE; Haves.
C_ (1988b). Effects of ambient ozone on respiratory function in healthy adults exercising outdoors. Am Rev
Respir Dis 138: 821-828.
Stafoggia. M; Forastiere. F; Faustini. A; Biggeri. A; Bisanti. L; Cadum. E; Cernigliaro. A; Mallone. S;
Pandolfi. P; Serinelli. M; Tessari. R; Vigotti. MA; Perucci. CA. (2010). Susceptibility factors to ozone-
related mortality: A population-based case-crossover analysis. Am J Respir Crit Care Med 182: 376-384.
http://dx.doi.org/10.1164/rccm.200908-1269OC
Steinvil. A; Fireman. E; Kordova-Biezuner. L; Cohen. M; Shapira. I; Berliner. S; Rogowski. O. (2009).
Environmental air pollution has decremental effects on pulmonary function test parameters up to one week
after exposure. Am J Med Sci 338: 273-279. http://dx.doi.org/10.1097/MAJ.Ob013e3181adb3ed
Steinvil. A; Kordova-Biezuner. L; Shapira. I; Berliner. S; Rogowski. O. (2008). Short-term exposure to air
pollution and inflammation-sensitive biomarkers. Environ Res 106: 51-61.
http://dx.doi.0rg/10.1016/i.envres.2007.08.006
Stenfors. N; Bosson. J; Helledav. R; Behndig. AF; Pourazar. J; Tornqvist. H; Kelly. FJ; Frew. AJ; Sandstrom.
T; Mudwav. IS; Blomber. A. (2010). Ozone exposure enhances mast-cell inflammation in asthmatic
airways despite inhaled corticosteroid therapy. Inhal Toxicol 22: 133-139.
http://dx.doi.org/10.3109/08958370903005736
Stenfors. N; Pourazar. J; Blomberg. A; Krishna. MT; Mudwav. I; Helledav. R; Kelly. FJ; Frew. AJ; Sandstrom.
X, (2002). Effect of ozone on bronchial mucosal inflammation in asthmatic and healthy subjects. Respir
Med 96: 352-358.
Stieb. DM; Szyszkowicz. M; Rowe. BH; Leech. JA. (2009). Air pollution and emergency department visits for
cardiac and respiratory conditions: A multi-city time-series analysis. Environ Health Global Access Sci
Source 8: 25. http://dx.doi.org/10.1186/1476-069X-8-25
6-292
-------
Strickland. MJ; Darrow. LA; Klein. M; Flanders. WD; Sarnat. JA; Waller. LA; Sarnat. SE; Mulholland. JA;
Tolbert. PE. (2010). Short-term associations between ambient air pollutants and pediatric asthma
emergency department visits. Am J Respir Crit Care Med 182: 307-316.
http://dx.doi.org/10.1164/rccm.200908-1201OC
Strickland. MJ: Darrow. LA; Mulholland. JA: Klein. M; Flanders. WD: Winquist. A; Tolbert. PE. (2011).
Implications of different approaches for characterizing ambient air pollutant concentrations within the
urban airshed for time-series studies and health benefits analyses. Environ Health Global Access Sci Source
10: 36. http://dx.doi.org/10.1186/1476-069X-10-36
Stylianou. M; Nicolich. MJ. (2009). Cumulative effects and threshold levels in air pollution mortality: Data
analysis of nine large US cities using the NMMAPS dataset. Environ Pollut 157: 2216-2223.
http://dx.doi.0rg/10.1016/i.envpol.2009.04.011
Symons. JM; Wang. L; Guallar. E; Howell. E; Dominici. F; Schwab. M; Ange. BA; Samet J: Ondov. J:
Harrison. D; Geyh. A. (2006). A case-crossover study of fine paniculate matter air pollution and onset of
congestive heart failure symptom exacerbation leading to hospitalization. Am J Epidemiol 164: 421-433.
http://dx.doi.org/10.1093/aie/kwj206
Szyszkowicz. M. (2008). Ambient air pollution and daily emergency department visits for ischemic stroke in
Edmonton, Canada. Int J Occup Med Environ Health 21: 295-300. http://dx.doi.org/10.2478/vl0001-008-
0029-5
Takeuchi. C: Galve. R; Nieva. J: Witter. DP; Wentworth. AD; Troseth. RP; Lerner. RA; Wentworth. P. Jr.
(2006). Proatherogenic effects of the cholesterol ozonolysis products, atheronal-A and atheronal-B.
Biochemistry 45: 7162-7170. http://dx.doi.org/10.1021M0604330
Tamer. L; Calikoglu. M; Ates. NA; Yildirim. H; Ercan. B; Saritas. E; Unhu A; Atik. U. (2004). Glutathione-S-
transferase gene polymorphisms (GSTT1, GSTM1, GSTP1) as increased risk factors for asthma.
Respirology 9: 493-498. http://dx.doi.Org/10.llll/j.1440-1843.2004.00657.x
Tankerslev. CG: Peng. RD; Bedga. D; Gabrielson. K; Champion. HC. (2010). Variation in echocardiographic
and cardiac hemodynamic effects of PM and ozone inhalation exposure in strains related to Nppa and Nprl
gene knock-out mice. Inhal Toxicol 22: 695-707. http://dx.doi.org/10.3109/08958378.2010.487549
Tepper. JL; Weiss. B; Cox. C. (1982). Microanalysis of ozone depression of motor activity. Toxicol Appl
Pharmacol64:317-326.
Tepper. JL; Weiss. B; Wood. RW. (1983). Behavioral indices of ozone exposure. In SD Lee; MG Mustafa; MA
Mehlman (Eds.), International symposium on the biomedical effects of ozone and related photochemical
oxidants (pp. 515526). Princeton NJ: Princeton Scientific.
Tepper. JS; Costa. PL; Lehmann. JR; Weber. MF; Hatch. GE. (1989). Unattenuated structural and biochemical
alterations in the rat lung during functional adaptation to ozone. Am J Respir Crit Care Med 140: 493-501.
Tepper. JS; Weiss. B; Wood. RW. (1985). Alterations in behavior produced by inhaled ozone or ammonia.
Toxicol Sci 5: 1110-1118.
Thaller. El; Petronella. SA; Hochman. D; Howard. S; Chhikara. RS; Brooks. EG. (2008). Moderate increases
in ambient PM2.5 and ozone are associated with lung function decreases in beach lifeguards. J Occup
Environ Med 50: 202-211. http://dx.doi.org/10.1097/JOM.Ob013e31816386b4
Thompson. AM; Zanobetti. A; Silverman. F; Schwartz. J; Coull. B; Urch. B; Speck. M; Brook. JR; Manno. M;
Gold. DR. (2010). Baseline Repeated Measures from Controlled Human Exposure Studies: Associations
between Ambient Air Pollution Exposure and the Systemic Inflammatory Biomarkers IL-6 and Fibrinogen.
Environ Health Perspect 118: 120-124. http://dx.doi.org/10.1289/ehp.0900550
Thomson. E; Kumarathasan. P; Goegan. P; Aubin. RA; Vincent. R. (2005). Differential regulation of the lung
endothelin system by urban particulate matter and ozone. Toxicol Sci 88: 103-113.
http://dx.doi.org/10.1093/toxsci/kFi272
Thomson. E; Kumarathasan. P; Vincent. R. (2006). Pulmonary expression of preproET-1 and preproET-3
mRNAs is altered reciprocally in rats after inhalation of air pollutants. Exp Biol Med 231: 979-984.
6-293
-------
Thomson. EM; Kumarathasan. P; Calderon-Garciduenas. L; Vincent. R. (2007). Air pollution alters brain and
pituitary endothelin-1 and inducible nitric oxide synthase gene expression. Environ Res 105: 224-233.
http://dx.doi.0rg/10.1016/i.envres.2007.06.005
Thurston. GD; Lippmann. M; Scott. MB; Fine. JM. (1997). Summertime haze air pollution and children with
asthma. Am J Respir Crit Care Med 155: 654-660.
Tolbert. PE; Klein. M; Peel. JL; Sarnat SE; Sarnat. JA. (2007). Multipollutant modeling issues in a study of
ambient air quality and emergency department visits in Atlanta. J Expo Sci Environ Epidemiol 17: S29-
S35. http://dx.doi.org/10.1038/si.ies.750Q625
Torres. A; Utell. MJ; Morow. PE; Voter. KZ; Whitin. JC: Cox. C: Loonev. RJ: Speers. DM; Tsai. Y; Frampton.
MW. (1997). Airway inflammation in smokers and nonsmokers with varying responsiveness to ozone. Am
J Respir Crit Care Med 156: 728-736.
Trenga. CA; Koenig. JQ; Williams. PV. (2001). Dietary antioxidants and ozone-induced bronchial
hyperresponsiveness in adults with asthma. Arch Environ Occup Health 56: 242-249.
Triche. EW: Gent. JF; Holford. TR; Belanger. K; Bracken. MB; Beckett. WS: Naeher. L; McSharry. JE;
Leaderer. BP (2006). Low-level ozone exposure and respiratory symptoms in infants. Environ Health
Perspect 114: 911-916. http://dx.doi.org/10.1289/ehp.8559
Turner. RM; Muscatello. DJ: Zheng. W: Willmore. A; Arendts. G. (2007). An outbreak of cardiovascular
syndromes requiring urgent medical treatment and its association with environmental factors: an ecological
study. Environ Health 6: 37. http://dx.doi.org/10.1186/1476-069X-6-37
Tyler. WS: Tyler. NK; Last. JA; Gillespie. MJ: Barstow. TJ. (1988). Comparison of daily and seasonal
exposures of young monkeys to ozone. Toxicology 50: 131-144.
U.S. EPA (U.S. Environmental Protection Agency). (1986). Air quality criteria for ozone and other
photochemical oxidants [EPA Report]. (EPA-600/8-84-020aF - EPA-600/8-84-020eF). Research Triangle
Park, NC. http://www.ntis.gov/search/product.aspx?ABBR=PB87142949
U.S. EPA (U.S. Environmental Protection Agency). (1996a). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/P-93/004AF). Research Triangle Park,NC.
U.S. EPA (U.S. Environmental Protection Agency). (1996f). Table 6-5. Lung inflammation and permeability
changes associated with ozone exposure [EPA Report]. In Air quality criteria for ozone and related
photochemical oxidants (pp. 22-27). (EPA/600/P-93/004AF). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (1996g). Table 6-6. Effects of ozone on host defense
mechanisms: Physical clearance [EPA Report]. In Air quality criteria for ozone and related photochemical
oxidants (pp. 38). (EPA/600/P-93/004AF). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (1996h). Table 6-7. Effects of ozone on host defense
mechanisms: Macrophage alterations [EPA Report]. In Air quality criteria for ozone and related
photochemical oxidants (pp. 40-41). (EPA/600/P-93/004AF). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (1996i). Table 6-8. Effects of ozone on host defense
mechanisms: Immunology [EPA Report]. In Air quality criteria for ozone and related photochemical
oxidants (pp. 45-46). (EPA/600/P-93/004AF). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (1996J). Table 6-9. Effects of ozone on host defense
mechanisms: Interactions with infectious agents [EPA Report]. In Air quality criteria for ozone and related
photochemical oxidants (pp. 50). (EPA/600/P-93/004AF). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (1996k). Table 6-10. Effects of ozone on conducting
airways [EPA Report]. In Air quality criteria for ozone and related photochemical oxidants (pp. 58-60).
(EPA/600/P-93/004AF). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (19961). Table 6-11. Effects of ozone on lung structure:
Short-term exposures (<2 weeks) [EPA Report]. In Air quality criteria for ozone and related photochemical
oxidants (pp. 61-63). (EPA/600/P-93/004AF). Research Triangle Park, NC.
6-294
-------
U.S. EPA (U.S. Environmental Protection Agency). (1996m). Table 6-13. Effects of ozone on pulmonary
function [EPA Report]. In Air quality criteria for ozone and related photochemical oxidants (pp. 86-87).
(EPA/600/P-93/004AF). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (1996n). Table 6-14. Effects of ozone on airway reactivity
[EPA Report]. In Air quality criteria for ozone and related photochemical oxidants (pp. 88-90). (EPA/600/P-
93/004AF). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (2006b). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/R-05/004AF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
U.S. EPA (U.S. Environmental Protection Agency). (2006d). Table AX5-7. Effects of ozone on lung host
defenses [EPA Report]. In Air quality criteria for ozone and related photochemical oxidants (pp. 8-16).
(EPA/600/R-05/004AF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
U.S. EPA (U.S. Environmental Protection Agency). (2006e). Table AX5-8. Effects of ozone on lung
permeability and inflammation [EPA Report]. In Air quality criteria for ozone and related photochemical
oxidants (pp. 17-26). (EPA/600/R-05/004AF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
U.S. EPA (U.S. Environmental Protection Agency). (2006f). Table AX5-9. Effects of ozone on lung structure:
Acute and subchronic exposures [EPA Report]. In Air quality criteria for ozone and related photochemical
oxidants (pp. 27-30). (EPA/600/R-05/004AF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
U.S. EPA (U.S. Environmental Protection Agency). (2006g). Table AX5-11. Effects of ozone on pulmonary
function [EPA Report]. In Air quality criteria for ozone and related photochemical oxidants (pp. 34-35).
(EPA/600/R-05/004AF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
U.S. EPA (U.S. Environmental Protection Agency). (2006h). Table AX5-12. Effects of ozone on airway
responsiveness [EPA Report]. In Air quality criteria for ozone and related photochemical oxidants (pp. 36-
42). (EPA/600/R-05/004AF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
Uchiyama. I; Simomura. Y; Yokoyama. E. (1986). Effects of acute exposure to ozone on heart rate and blood
pressure of the conscious rat. Environ Res 41: 529-537.
Uchiyama. I; Yokoyama. E. (1989). Effects of short- and long-term exposure to ozone on heart rate and blood
pressure of emphysematous rats. Environ Res 48: 76-86. http://dx.doi.org/10.1016/80013-9351(89)80087-8
Ulmer. C: Kopp. M; Ihorst. G: Frischer. T; Forster. J; Kuehr. J. (1997). Effects of ambient ozone exposures
during the spring and summer of 1994 on pulmonary function of schoolchildren. Pediatr Pulmonol 23: 344-
353. http://dx.doi.org/10.1002/(SICI)1099-0496(199705)23:5<344::AID-PPUL6>3.0.CO:2-K
Ultman. JS: Ben-Jebria. A; Arnold. SF. (2004). Uptake distribution of ozone in human lungs: Intersubject
variability in physiologic response (pp. 1-23; discussion 25-30). (ISSN 1041-5505
HEI Research Report 125). Boston, MA: Health Effects Institute. http://pubs.healtheffects.org/view.php?id=70
Urch. B; Silverman. F; Corey. P; Brook. JR; Lukic. KZ; Raiagopalan. S: Brook. RD. (2005). Acute blood
pressure responses in healthy adults during controlled air pollution exposures. Environ Health Perspect 113:
1052-1055.
Urch. B; Speck. M; Corey. P; Wasserstein. D; Manno. M; Lukic. KZ; Brook. JR; Liu. L; Coull. B; Schwartz. J:
Gold. PR; Silverman. F. (2010). Concentrated ambient fine particles and not ozone induce a systemic
interleukin-6 response in humans. Inhal Toxicol 22: 210-218.
http://dx.doi.org/10.3109/08958370903173666
6-295
-------
Vagaggini. B; Bartoli. MLE; Cianchetti. S: Costa. F; Bacci. E; Dente. FL; Di Franco. A; Malagrino. L;
Paggiaro. P. (2010). Increase in markers of airway inflammation after ozone exposure can be observed also
in stable treated asthmatics with minimal functional response to ozone. Respir Res 11:5.
http://dx.doi.org/10.1186/1465-9921-ll-5
Vagaggini. B; Cianchetti. S: Bartoli. M; Ricci. M; Bacci. E; Dente. FL; Di Franco. A; Paggiaro. P. (2007).
Prednisone blunts airway neutrophilic inflammatory response due to ozone exposure in asthmatic subjects.
Respiration 74: 61-58. http://dx.doi.org/10.1159/000096078
Vagaggini. B; Taccola. M; Cianchetti. S: Carnevali. S: Bartoli. ML; Bacci. E; Dente. FL; Di Franco. A;
Giannini. D; Paggiaro. PL. (2002). Ozone exposure increases eosinophilic airway response induced by
previous allergen challenge. Am J Respir Crit Care Med 166: 1073-1077.
Vagaggini. B; Taccola. M; Conti. I; Carnevali. S: Cianchetti. S: Bartoli. ML; Bacci. E; Dente. FL; Di Franco.
A; Giannini. D; Paggiaro. PL. (2001). Budesonide reduces neutrophilic but not functional airway response
to ozone in mild asthmatics. Am J Respir Crit Care Med 164: 2172-2176.
Valacchi. G: Pecorelli. A; Mencarelli. M; Maioli. E; Davis. PA. (2009). Beta-carotene prevents ozone-induced
proinflammatory markers in murine skin. Toxicol Ind Health 25: 241-247.
http://dx.doi.org/10.1177/0748233709103030
Van Loveren. H; Krajnc. El; Rombout. PJ: Blommaert. FA; Vos. JG. (1990). Effects of ozone,
hexachlorobenzene, and bis(tri-n-butyltin)oxide on natural killer activity in the rat lung. Toxicol Appl
Pharmacol 102: 21-33.
Van Loveren. H; Rombout. PJA; Wagenaar. SS: Walvoort. HC: Vos. JG. (1988). Effects of ozone on the
defense to a respiratory Listeria monocytogenes infection in the rat: Suppression of macrophage function
and cellular immunity and aggravation of histopathology in lung and liver during infection. Toxicol Appl
Pharmacol 94: 374-393.
Vanguilder. HP; Freeman. WM. (2011). The hippocampal neuroproteome with aging and cognitive decline:
Past progress and future directions. Front Aging Neurosci 3: 8. http://dx.doi.org/10.3389/fnagi.2011.00008
Vesely. PL; Giordano. AT; Raska-Emerv. P; Montgomery. MR. (1994a). Increase in atrial natriuretic factor in
the lungs, heart, and circulatory system owing to ozone. Chest 105: 1551-1554.
Vesely. PL; Giordano. AT; Raska-Emerv. P; Montgomery. MR. (1994b). Ozone increases amino- and carboxy-
terminal atrial natriuretic factor prohormone peptides in lung, heart, and circulation. J Biochem Mol
Toxicol 9: 107-112. http://dx.doi.org/10.1002/ibt.2570090208
Vesely. PL; Giordano. AT; Raska-Emerv. P; Montgomery. MR. (1994c). Ozone increases atrial natriuretic
peptides in heart, lung and circulation of aged vs adult animals. Gerontology 40: 227-236.
http://dx.doi.org/10.1159/000213590
Villeneuve. PJ: Chen. L; Rowe. BH; Coates. F. (2007). Outdoor air pollution and emergency department visits
for asthma among children and adults: A case-crossover study in northern Alberta, Canada. Environ Health
Global Access Sci Source 6: 40. http://dx.doi.org/10.1186/1476-069X-6-40
Villeneuve. PJ: Chen. L; Stieb. D: Rowe. BH. (2006a). Associations between outdoor air pollution and
emergency department visits for stroke in Edmonton, Canada. Eur J Epidemiol 21: 689-700.
http://dx.doi.org/10.1007/sl0654-006-9050-9
Villeneuve. PJ: Doiron. MS: Stieb. D: Dales. R; Burnett. RT; Dugandzic. R. (2006b). Is outdoor air pollution
associated with physician visits for allergic rhinitis among the elderly in Toronto, Canada? Allergy 61: 750-
758. http://dx.doi.0rg/10.llll/i.1398-9995.2006.01070.x
Vincent. R: Janzen. EG: Chen. G: Kumarathasan. P; Haire. PL; Guenette. J: Chen. JZ; Bray. TM. (1996a).
Spin trapping study in the lungs and liver of F344 rats after exposure to ozone. Free Radic Res 25: 475-488.
6-296
-------
Von Klot. S: Peters. A; Aalto. P; Bellander. T; Berglind. N; D'lppoliU D; Elosua. R; Hormann. A; Kulmala. M;
Lanki. T; Lowel H; Pekkanen. J; Picciotto. S: Sunyer. J; Forastiere. F; Group. HEoPoSSHS. (2005).
Ambient air pollution is associated with increased risk of hospital cardiac readmissions of myocardial
infarction survivors in five European cities. Circulation 112: 3073-3079.
http://dx.doi.org/10.1161/CIRCULATIONAHA.105.548743
Vovnow. JA; Fischer. BM; Zheng. S: Potts. EN; Grover. AR; Jaiswal. AK; Ohio. AJ; Foster. WM. (2009).
NAD(P)H quinone oxidoreductase 1 is essential for ozone-induced oxidative stress in mice and humans.
Am J Respir Cell Mol Biol 41: 107-113. http://dx.doi.org/10.1165/rcmb.2008-0381OC
Wagner. JG: Harkema. JR; Jiang. Q; Illek. B; Ames. BN; Peden. DB. (2009). Gamma-tocopherol attenuates
ozone-induced exacerbation of allergic rhinosinusitis in rats. Toxicol Pathol 37: 481-491.
http://dx.doi.org/10.1177/0192623309335630
Wagner. JG: Jiang. Q; Harkema. JR: Illek. B; Patel. DP; Ames. BN; Peden. DB. (2007). Ozone enhancement
of lower airway allergic inflammation is prevented by gamma-tocopherol. Free Radic Biol Med 43: 1176-
1188. http://dx.doi.0rg/10.1016/i.freeradbiomed.2007.07.013
Ward. DJ: Roberts. KT; Jones. N; Harrison. RM; Ayres. JG: Hussain. S: Walters. S. (2002). Effects of daily
variation in outdoor particulates and ambient acid species in normal and asthmatic children. Thorax 57:
489-502. http://dx.doi.0rg/10.1136/thorax.57.6.489
Watkinson. WP; Aileru. AA; Dowd. SM; Doerfler. PL; Tepper. JS: Costa. PL. (1993). Acute effects of ozone
on heart rate and body temperature in the unanesthetized, unrestrained rat maintained at different ambient
temperatures. Inhal Toxicol 5: 129-147.
Watkinson. WP: Campen. MJ; Wichers. LB; Nolan. JP; Costa. PL. (2003). Cardiac and thermoregulatory
responses to inhaled pollutants in healthy and compromised rodents: Modulation via interaction with
environmental factors. Environ Res 92: 35-47.
Weinmann. GG: Bowes. SM; Gerbase. MW: Kimball. AW: Frank. R. (1995a). Response to acute ozone
exposure in healthy men. Results of a screening procedure. Am J Respir Crit Care Med 151: 33-40.
Weinmann. GG: Weidenbach-Gerbase. M; Foster. WM; Zacur. H; Frank. R. (1995c). Evidence for ozone-
induced small-airway dysfunction: Lack of menstrual-cycle and gender effects. Am J Respir Crit Care Med
152: 988-996.
Wellenius. GA; Bateson. TF; Mittleman. MA; Schwartz. J. (2005). Paniculate air pollution and the rate of
hospitalization for congestive heart failure among medicare beneficiaries in Pittsburgh, Pennsylvania. Am J
Epidemioll61: 1030-1036.
Wellenius. GA: Yeh. GY: Coull. BA: Suh. HH: Phillips. RS: Mittleman. MA. (2007). Effects of ambient air
pollution on functional status in patients with chronic congestive heart failure: A repeated-measures study.
Environ Health 6: 1-7. http://dx.doi.org/10.1186/1476-069X-6-26
Wentworth. P. Jr; Nieva. J: Takeuchi. C: Galve. R: Wentworth. AP; Pillev. RB; Pelaria. GA: Saven. A; Babior.
BM; Janda. KP; Eschenmoser. A; Lerner. RA. (2003). Evidence for ozone formation in human
atherosclerotic arteries. Science 302: 1053-1056. http://dx.doi.org/10.1126/science.1089525
Wheeler. A; Zanobetti. A; Gold. PR; Schwartz. J: Stone. P; Suh. HH. (2006). The relationship between
ambient air pollution and heart rate variability differs for individuals with heart and pulmonary disease.
Environ Health Perspect 114: 560-566.
Wiester. MJ: Watkinson. WP: Costa. PL; Crissman. KM; Richards. JH; Winsett PW: Highfill. JW. (1996c).
Ozone toxicity in the rat. III. Effect of changes in ambient temperature on pulmonary parameters. J Appl
PhysiolSl: 1691-1700.
Wiwatanadate. P: Liwsrisakun. C. (2011). Acute effects of air pollution on peak expiratory flow rates and
symptoms among asthmatic patients in Chiang Mai, Thailand. Int J Hyg Environ Health 214: 251-257.
http://dx.doi.0rg/10.1016/i.iiheh.2011.03.003
6-297
-------
Wiwatanadate. P; Trakultivakorn. M. (2010). Air pollution-related peak expiratory flow rates among asthmatic
children in Chiang Mai, Thailand. Inhal Toxicol 22: 301-308.
http://dx.doi.org/10.3109/08958370903300327
Wong. CM; Ma. S: AJ. H; Lam. TH. (1999a). Does ozone have any effect on daily hospital admissions for
circulatory diseases? J Epidemiol Community Health 53: 580-581.
Wong. CM; Vichit-Vadakan. N; Vajanapoom. N; Ostro. B; Thach. TQ; Chau. PY; Chan. EK; Chung. RY; Ou.
CO: Yang. L; Peiris. JS: Thomas. GN; Lam. TH; Wong. TW: Hedlev. AJ: Kan. H; Chen. B; Zhao. N;
London. SJ: Song. G: Chen. G: Zhang. Y; Jiang. L; Qian. Z; He. Q: Lin. HM; Kong. L; Zhou. D; Liang. S:
Zhu. Z: Liao. D: Liu. W: Bentlev. CM: Dan. J: Wang. B: Yang. N: Xu. S: Gong. J: Wei. H: Sun. H: Oin. Z.
(2010). Part 5. Public health and air pollution in Asia (PAPA): A combined analysis of four studies of air
pollution and mortality. In Public Health and Air Pollution in Asia (PAPA): Coordinated Studies of Short-
Term Exposure to Air Pollution and Daily Mortality in Four Cities (pp. 377-418). Boston, MA: Health
Effects Institute. http://pubs.healtheffects.org/view.php?id=348
Wong. CM; Yang. L; Thach. TQ: Chau. PY; Chan. KP; Thomas. GN; Lam. TH; Wong. TW: Hedlev. AJ: Peiris.
JS. (2009). Modification by influenza on health effects of air pollution in Hong Kong. Environ Health
Perspect 117: 248-253. http://dx.doi.org/10.1289/ehp.11605
Wong. TW: Lau. TS: Yu. TS: Neller. A; Wong. SL; Tarn. W: Pang. SW. (1999b). Air pollution and hospital
admissions for respiratory and cardiovascular diseases in Hong Kong. Occup Environ Med 56: 679-683.
Wu. CF; Kuo. 1C: Su. TC: Li. YR; Lin. LY; Chan. CC: Hsu. SC. (2010). Effects of personal exposure to
paniculate matter and ozone on arterial stiffness and heart rate variability in healthy adults. Am J Epidemiol
171: 1299-1309. http://dx.doi.org/10.1093/aje/kwq060
Xia. Y; Tong. H. (2006). Cumulative effects of air pollution on public health. Stat Med 25: 3548-3559.
http://dx.doi.org/10.1002/sim.2446
Yang. CY. (2008). Air pollution and hospital admissions for congestive heart failure in a subtropical city:
Taipei, Taiwan. J Toxicol Environ Health A 71: 1085-1090. http://dx.doi.org/10.1080/15287390802114428
Yang. CY; Chen. YS: Yang. CH; Ho. SC. (2004). Relationship between ambient air pollution and hospital
admissions for cardiovascular diseases in Kaohsiung, Taiwan. J Toxicol Environ Health A 67: 483-493.
Yang. Q; Chen. Y; Krewski. D; Burnett. RT; Shi. Y; Mcgrail. KM. (2005b). Effect of short-term exposure to
low levels of gaseous pollutants on chronic obstructive pulmonary disease hospitalizations. Environ Res
99: 99-105. http://dx.doi.Org/10.1016/i.envres.2004.09.014
Yokoyama. E; Uchiyama. I; Arito. H. (1989). Extrapulmonary effects of low level ozone exposure. In T
Schneider; SD Lee; GJR Wolters; LD Grant (Eds.), Atmospheric ozone research and its policy implications:
proceedings of the 3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands (pp.
301-309). Amsterdam, The Netherlands: Elsevier.
Yoon. HK; Cho. HY; Kleeberger. SR. (2007). Protective role of matrix metalloproteinase-9 in ozone-induced
airway inflammation. Environ Health Perspect 115: 1557-1563. http://dx.doi.org/10.1289/ehp.10289
Zanobetti. A; Canner. MJ; Stone. PH; Schwartz. J; Sher. D; Eagan-Bengston. E; Gates. KA; Hartley. LH; Suh.
H; Gold. DR. (2004). Ambient pollution and blood pressure in cardiac rehabilitation patients. Circulation
110: 2184-2189. http://dx.doi.org/10.1161/01.cir.0000143831.33243.d8
Zanobetti. A; Gold. PR; Stone. PH; Suh. HH; Schwartz. J; Coull. BA; Speizer. FE. (2010). Reduction in heart
rate variability with traffic and air pollution in patients with coronary artery disease. Environ Health
Perspect 118: 324-330. http://dx.doi.org/10.1289/ehp.0901003
Zanobetti. A; Schwartz. J. (2006). Air pollution and emergency admissions in Boston, MA. J Epidemiol
Community Health 60: 890-895. http://dx.doi.org/10.1136/iech.2005.039834
Zanobetti. A; Schwartz. J. (2008a). Is there adaptation in the ozone mortality relationship: A multi-city case-
crossover analysis. Environ Health 7: 22. http://dx.doi.org/10.1186/1476-069X-7-22
6-298
-------
Zanobetti. A; Schwartz. J. (2008b). Mortality displacement in the association of ozone with mortality:
An analysis of 48 cities in the United States. Am J Respir Crit Care Med 177: 184-189.
http://dx.doi.org/10.1164/rccm.200706-823OC
6-299
-------
INTEGRATED HEALTH EFFECTS OF LONG-TERM OZONE
EXPOSURE
7.1 Introduction
This chapter reviews, summarizes, and integrates the evidence on relationships
between health effects and long-term exposures to O3. Both epidemiologic and
toxicological studies provide a basis for examining long-term O3 exposure health
effects for respiratory effects, cardiovascular effects, reproductive and developmental
effects, central nervous system effects, cancer outcomes, and mortality. Long-term
exposure has been defined as a duration of approximately 30 days (1 month) or
longer1. However, in order to characterize the weight of evidence for the effects of
O3 on reproductive and developmental effects in a consistent, cohesive and
integrated manner, results from both short-term and long-term exposure periods are
included in that section, and are identified accordingly in the text and tables.
Conclusions from the 2006 O3 AQCD (U.S. EPA. 2006b) are summarized briefly at
the beginning of each section, and the evaluation of evidence from recent studies
builds upon what was available during the previous review. For each health outcome
(e.g., respiratory disease, lung function), results are summarized for studies from the
specific scientific discipline, i.e., epidemiologic and toxicological studies. The major
sections (i.e., respiratory, cardiovascular, mortality, reproductive/developmental,
cancer) conclude with summaries of the evidence for the various health outcomes
within that category and integration of the findings that lead to conclusions regarding
causality based upon the framework described in the Preamble to this ISA.
Determination of causality is made for the overall health effect category, such as
respiratory effects, with coherence and plausibility being based on evidence from
across disciplines and also across the suite of related health outcomes, including
cause-specific mortality.
As mentioned in Chapter 2 (Section 2.3), epidemiologic studies generally present O3-
related effect estimates for mortality and morbidity health outcomes based on an
incremental change in exposure. Studies traditionally present the relative risk per an
incremental change equal to the interquartile range in O3 concentrations or some
other arbitrary value (e.g., 10 ppb). Additionally, various exposure metrics are used
in O3 epidemiologic studies, with the three most common being the maximum 1-h
average within a 24-hour period (1-h max), the maximum 8-h average within a
24-hour period (8-h max), and 24-h average (24-h avg). For the purpose of
presenting results from studies that use different exposure metrics, EPA consistently
applies the same O3 increments to facilitate comparisons between the results of
various studies that may present results for different incremental changes.
Differences due to the use of varying exposure metrics (e.g., 1-h max, 24-h avg)
1 Unless otherwise specified, the term "chronic" generally refers to an annual exposure duration for epidemiology studies and a
duration of greater than 10% of the lifespan of the animal in toxicological studies.
7-1
-------
become less apparent when averaged across longer exposure periods, because levels
are typically lower and less variable. As such, throughout this chapter an increment
of 10 ppb was consistently applied across studies, regardless of exposure metric, to
facilitate comparisons between the results from these studies.
7.2 Respiratory Effects
Studies reviewed in the 2006 O3 AQCD examined evidence for relationships
between long-term O3 exposure (several months to yearly) and effects on respiratory
health outcomes including declines in lung function, increases in inflammation, and
development of asthma in children and adults. Animal toxicology data provided a
clearer picture indicating that long-term O3 exposure may have lasting effects.
Chronic exposure studies in animals have reported biochemical and morphological
changes suggestive of irreversible long-term O3 impacts on the lung. In contrast to
supportive evidence from chronic animal studies, the epidemiologic studies on
longer-term (annual) lung function declines, inflammation, and new asthma
development remained inconclusive.
Several studies reviewed in the 2006 O3 AQCD (Horak et al., 2002; Frischer et al,
1999) collectively indicated that O3 exposure averaged over several summer months
was associated with smaller increases in lung function growth in children. For longer
averaging periods (annual), the definitive analysis in the Children's Health Study
(CHS) reported by Gauderman et al. (2004) provided little evidence that such long-
term exposure to ambient O3 was associated with significant deficits in the growth
rate of lung function in children in contrast to the effects observed with other
pollutants such as acid vapor, NO2, and PM2.5. Limited epidemiologic research
examined the relationship between long-term O3 exposures and inflammation.
Consistent with evidence of inflammation and allergic responses reported in
experimental studies, an association between 30-day average O3 and increased
eosinophil levels was observed in an Austrian study (Frischer et al.. 2001).
The cross-sectional studies available for the 2006 O3 AQCD detected no associations
between long-term O3 exposures and asthma prevalence, asthma-related symptoms
or allergy to common aeroallergens in children after controlling for covariates.
However, longitudinal studies provided evidence that long-term O3 exposure
influences the risk of asthma development in children (McConnell et al.. 2002) and
adults (McDonnell et al.. 1999a: Greeretal. 1993).
New evidence presented below reports interactions between genetic variants and
long-term O3 exposure in effects on new onset asthma in U.S. cohorts in multi-
community studies where protection by specific oxidant gene variants was restricted
to children living in low O3 communities. Related studies report coherent
relationships between respiratory symptoms among asthmatics and long-term O3
exposure. This evidence for respiratory effects associated with long-term O3
exposure is supported by a large evidence base indicating associations of short-term
exposure to O3 with increases in respiratory symptoms and asthma medication use in
7-2
-------
children with asthma (Section 6.2.4.1) and asthma hospitalizations in children
(Section 6.2.7.2). A new line of evidence reports a positive concentration-response
relationship between first asthma hospitalization and long-term O3 exposure. Results
from recent studies examining pulmonary function, inflammation, and allergic
responses are also presented.
7.2.1 Asthma
7.2.1.1 New Onset Asthma
Asthma is a heterogeneous disease with a high degree of temporal variability. Its
progression and symptoms can vary within an individual's experience over time.
The course of asthma may vary markedly between young children, older children and
adolescents, and adults. This variation is probably more dependent on age than on
symptoms (NHLBI, 2007). Longitudinal cohort studies have examined associations
between long-term O3 exposures and the onset of asthma in adults and children
(McConnell et al., 2002; McDonnell et al., 1999a; Greer etal, 1993), with results
indicating a direct effect of long-term O3 exposure on asthma risk in adults and effect
modification by O3 in children.
Associations between long-term O3 exposure and new cases of asthma were reported
in a cohort of nonsmoking adults in California (McDonnell et al., 1999a; Greer et al.,
1993). The Adventist Health and Smog (AHSMOG) study cohort of 3,914 (age 27 to
87 years, 36% male) was drawn from nonsmoking, non-Hispanic white California
Seventh Day Adventists, who were surveyed in 1977, 1987, and 1992. To be eligible,
subjects had to have lived 10 or more years within 5 miles of their current residence
in 1977. Residences from 1977 onward were followed and linked in time and space
to interpolate concentrations of O3, PM10, SO42", SO2, and NO2. New asthma cases
were defined as self-reported doctor-diagnosed asthma at either the 1987 or 1992
follow-up questionnaire among those who had not reported having asthma upon
enrollment in 1977. During the 10-year follow-up (1977 to 1987), the incidence of
new asthma was 2.1% for males and 2.2% for females (Greer et al., 1993). Ozone
concentration data were not provided. A relative risk of 3.12 (95% CI: 1.16, 5.85) per
10-ppb increase in annual mean O3 (exposure metric not stated) was observed in
males, compared to a relative risk of 0.94 (95% CI: 0.65, 1.34) in females.
In the 15-year follow-up study (1977-1992), 3.2% of the eligible males and a slightly
greater 4.3% of the eligible females developed adult asthma (McDonnell et al..
1999a). The mean 20-year average (1973-1992) for 8-h avg O3 (9 a.m. to 5 p.m.) was
46.5 ppb (SD 15.3). For males, the relative risk of developing asthma was 1.31
(95% CI: 1.01, 1.71) per 10-ppb increase in 8-h avg O3. Once again, there was no
evidence of a positive association between O3 and new-onset asthma in females
(relative risk of 0.94 [95% CI: 0.87, 1.02]). The lack of an association does not
necessarily indicate no effect of O3 on the development of asthma among females.
7-3
-------
For example, differences between females and males in time-activity patterns may
influence relative exposures to ambient O3. During summer 1992, the mean (SD)
hours per week spent outdoors for male and female asthma cases were 13.8 (10.6)
and 11.4 (10.9), respectively, indicating potential greater misclassification of
exposure in females. None of the other pollutants (PMi0, SO42", SO2, and NO2) were
associated with development of asthma in either males or females. Adjusting for
copollutants did not diminish the association between O3 and asthma incidence for
males. In no case was the O3 coefficient reduced by more than 10% in the two-
pollutant models compared to the model containing O3 alone. The consistency of the
results in the two studies with different follow-up times, as well as the independent
and robust association between annual mean O3 concentrations and asthma
incidence, provide supportive evidence that long-term O3 exposure may be
associated with the development of asthma in adult males. However, because the
AHSMOG cohort was drawn from a narrow subject definition, the representativeness
of this cohort to the general U.S. population may be limited.
In children, the relationship between long-term O3 exposure and new onset asthma
has been extensively investigated in the CHS. In this cohort, evidence provides
stronger support for long-term O3 exposure modifying the risk of new onset asthma
associated with other potential risk factors than having a main effect on new onset
asthma. Initiated in the early 1990s, the CHS was originally designed to examine
whether long-term exposure to ambient pollutants was related to chronic respiratory
outcomes in children in 12 communities in southern California (Peters et al., 1999b;
Peters et al., 1999a). New-onset asthma was classified as having no prior history of
asthma at study entry with subsequent report of physician-diagnosed asthma at
follow-up with the date of onset assigned to be the midpoint of the interval between
the interview date when asthma diagnosis was first reported and the previous
interview date. In a cohort recruited during 2002-2003 and followed for three years
beginning in kindergarten or first grade, McConnell et al. (2010) reported a hazard
ratio for new onset asthma of 0.76 (95% CI: 0.38, 1.54) comparing the communities
with the highest (59.8 ppb) and lowest (29.5 ppb) annual average of 8-h avg (10
a.m.-6 p.m.) O3. With adjustment for school and residential modeled non-freeway
traffic-related exposure, the estimated HR for O3 was 1.01 (95% CI: 0.49, 2.11).
Similarly in a cohort recruited in 1993, asthma risk was not higher for residents of
the six high-O3 communities versus residents of the six low-O3 communities
(McConnell et al., 2002). In this study, 3,535 initially nonasthmatic children (ages 9
to 16 years at enrollment) were followed for up to 5 years, during which 265 cases of
new-onset asthma were identified. Communities were stratified by 4-year average
1-h max O3 levels, with six high-O3 communities (mean 75.4 ppb [SD 6.8]) and six
low-O3 communities (mean 50.1 ppb [SD 11.0]). Within the high-O3 communities,
asthma risk was 3.3 (95% CI: 1.9, 5.8) times greater for children who played three or
more sports as compared with children who played no sports. None of the children
who lived in high-O3 communities and played three or more sports had a family
history of asthma. In models with individual sports entered as dummy variables, only
tennis was significantly associated with asthma and only in the high O3 communities.
This association was absent in the low-O3 communities (relative risk of 0.8 [95% CI:
7-4
-------
0.4, 1.6]). The overall observed pattern of effects of sports participation on asthma
risk was robust to adjustment for SES, history of allergy, family history of asthma,
insurance, maternal smoking, and BMI.
Analyses aimed at distinguishing the effects of O3 from effects of other pollutants
indicated that in communities with high O3 and low levels of other pollutants there
was a 4.2-fold (95% CI: 1.6, 10.7) increased risk of asthma in children playing three
or more sports, compared to children who played no sports. The relative risk in
children playing three or more sports was slightly lower (3.3 [95% CI: 1.6, 6.9]) in
communities with a combination of high levels of O3 and other pollutants. Ozone
concentrations were not strongly correlated with PMi0, PM2.5, NO2, or inorganic acid
vapors, and no associations with asthma were found for these other pollutants. These
results provide additional support that the effects of physical activity on asthma are
modified by long-term O3 exposure. Overall, the results from McConnell et al.
(2002) suggest that playing sports may indicate greater outdoor activity when O3
levels are higher and an increased ventilation rate, which may lead to increased O3
exposure. It should be noted, however, that these findings were based on a small
number of new asthma cases (n = 29 among children who played three or more
sports) and were not based on a priori hypotheses.
Recent studies from the CHS provide evidence for gene-environment interactions in
effects on new-onset asthma by indicating that the lower risks associated with
specific genetic variants are found in children who live in lower O3 communities
(Islam et al.. 2009: Islam et al.. 2008: Orvszczvn et al.. 2007: Lee et al.. 2004b:
Gilliland et al.. 2002). Risk for new-onset asthma is related in part to genetic
susceptibility, behavioral factors and environmental exposure (Gilliland et al.. 1999).
Gene-environment interactions in asthma have been well discussed in the literature
(von Mutius. 2009: Holgate et al.. 2007: Martinez. 2007a. b; Rahman et al.. 2006:
Hoffianetal..20Q5: Kleeberger and Peden. 2005: Ober. 2005). Complex chronic
diseases, such as asthma, are partially the result of a sequence of biochemical
reactions involving exposures to various environmental agents metabolized by a
number of different genes (Conti et al., 2003). Oxidative stress has been proposed to
underlie these mechanistic hypotheses (Gilliland et al., 1999). Genetic variants may
impact disease risk directly or modify disease risk by affecting internal dose of
pollutants and other environmental agents and/or their reaction products or by
altering cellular and molecular modes of action. Understanding the relation between
genetic polymorphisms and environmental exposure can help identify high-risk
subgroups in the population and provide better insight into pathway mechanisms for
these complex diseases.
CHS analyses have found that asthma risk is related to interactions between O3 and
variants in genes for enzymes such as heme-oxygenase (HO-1), arginases (ARG1
and 2), and glutathione S transferase PI (GSTP1) (Himes et al.. 2009: Islam et al..
2008: Li et al.. 2008: Hanene et al.. 2007: Ercan et al.. 2006: Li et al.. 2006a: Tamer
et al.. 2004: Gilliland et al.. 2002). Biological plausibility for these findings is
provided by evidence that these enzymes have antioxidant and/or anti-inflammatory
activity and participate in well recognized modes of action in asthma pathogenesis.
7-5
-------
Further, several lines of evidence demonstrate that secondary oxidation products of
O3 initiate the key modes of action that mediate downstream health effects
(Section 5.3.2). For example, HO-1 has been found to respond rapidly to oxidants,
have anti-inflammatory and anti-oxidant effects (Exner et al.. 2004), relax airway
smooth muscle, and be induced in airways during asthma (Carter et al., 2004).
The GSTP1 Val/Val genotype has been associated with increased risk of having
atopic asthma (Tamer et al.. 2004). Gene-environment interactions are discussed in
greater detail in Section 5.4.2.1.
Islam et al. (2008) found that functional polymorphisms of the heme oxygenase-1
gene (HMOX-1, [(GT)n repeat]) influenced the risk of new-onset asthma, depending
on ethnicity and long-term community O3 concentrations. Ozone-gene interactions
were not found for variants in other antioxidant genes: catalase (CAT [-262C >T
-844C >TO]) or and manganese superoxide dismutase (MNSOD, [Ala-9Val]).
Analyses were restricted to children of Hispanic (n = 576) or non-Hispanic white
ethnicity (n = 1,125) and were conducted with long-term pollutant levels averaged
from 1994 to 2003. The effect of ambient air pollution on the relationship between
genetic polymorphism and new-onset asthma was assessed using Cox proportional
hazard regression models where the community specific average air pollution levels
were fitted as a continuous variable together with the appropriate interaction terms
for genes and air pollutants and a random effect of community (Berhane et al., 2004).
Over the follow-up period, 160 new cases of asthma were diagnosed (Islam et al..
2008). For HMOX-1, the interaction (p = 0.003) indicated a greater protective effect
of the S-allele (short, <23 (GT)n repeats) compared to the L-allele (long, >23
repeats) among non-Hispanic white children who lived in the low O3 community
(nonparallelism presented in Figure 7-1). Among children residing in low-O3
communities, the hazard ratio (HR) of new onset asthma associated with the S-allele
was 0.44 (95% CI: 0.23, 0.83) compared to non-Hispanic white children who lived in
low O3 communities and had no S-alleles. Biological plausibility for these results is
provided by evidence that the S-allele variant of HMOX-1 is more readily induced
than those with more numerous repeats. The S-allele was found to have a less
protective effect in non-Hispanic white children who resided in high O3 communities
(HR = 0.88; [95% CI: 0.33, 2.34] compared to non-Hispanic white children in low
O3 communities with no S-allele). Because HMOX-1 variants were not associated
with asthma risk in Hispanic children, effect modification by O3 was not
investigated. No significant interactions were observed between PMi0 or other
pollutants and the HMOX-1 gene; quantitative results were not presented. Average
O3 levels showed low correlation with the other monitored pollutants. The authors
did not consider the lack of adjustment for multiple testing to be a concern in this
analysis because the selection of the genes was based on a priori hypotheses defined
by a well-studied biological pathway, in which oxidative stress serves as the link
among O3 exposure, enzyme activity, and asthma.
Collectively, results from Islam et al. (2008) indicate that a variant in HMOX-1 that
produces a more readily inducible enzyme is associated with lower risk of new-onset
asthma in children who live in low O3 communities. Results were not presented for
7-6
-------
the main effects relating new-onset asthma to O3 exposure. However, they do
indicate that that in environments of low ambient O3, enzymes with greater
antioxidative activity may have the capacity to counter any temporary imbalance in
an oxidant-antioxidant relationship. However, in the presence of high background
O3, the protective effect may be attenuated because with higher exposure to oxidants,
the antioxidant genes may be at their maximal level of inducibility, and variation in
promoters no longer affects levels of expression. Supporting evidence is provided by
Schroer et al. (2009). who found that infants with multiple environmental exposures
were at increased risk of wheeze regardless of variant in GSTP1, which encodes a
gene with antioxidant activity.
2.5
1.5
0.5
Interaction of Gene presence and Ozone Level on the
Hazard Ratio of New Onset Asthma (P-value of 0.003)
;rence
r
o
a:
0.441
(2.4
Children with no S-Allele
{0.83) _ —
i-" ~*~ ^Dhildren with S-Allele
}} (2.:
.0..24
5-4)
i 0.88
(0.28)
Low
(38.4 ppb)
Community Mean Ozone Level
High
(55.2 ppb)
(Confidence limits based on comparison with reference group)
Note: An interaction p-value of 0.003 was obtained from the hierarchical two stage Cox proportional hazard model fitting the
community specific O3 and controlling for random effect of the communities. The interaction indicates there is a greater protective
effect of having a heme-oxygenase S-allele compared to having the L-allele among children living in communities with lower long-
term ambient O3 concentrations. The HRs are off-set as opposed to overlapping in the figure to allow clearer presentation of the
results.
Source: Developed by EPA with data from Islam et al. (2008) (data used with permission of American Thoracic Society).
Figure 7-1 Interaction of heme-oxygenase genetic variants and Os level on the
Hazard Ratio (HR) of new-onset asthma in the 12 Children's Health
Study communities.
7-7
-------
Expanding on the results of McConnell et al. (2002), Islam et al. (2009) provided
evidence that variants in GSTM1 and GSTP1 may influence associations between
outdoor exercise and new onset asthma. A primary conclusion that the authors (Islam
et al.. 2009) reported was that the GSTP1 lie/lie and GSTM1 null genotypes
increased risk of new onset asthma during adolescence. The highest risk was found
for participation in three or more team sports (compared to no sports) among children
with GSTP1 lie/He genotype living in high-O3 communities (HR: 6.15, [95% CI: 2.2,
7.4]). No three-way interaction was found for GSTM1. These results demonstrate the
potential importance of a combination of genetic variability, O3 exposure, and
outdoor activity on asthma risk. It is important to note that while some studies have
found a modifying role of air pollution on the association between GSTP1 lie/He and
asthma in children (Lee et al.. 2004b). others have found that the GSTP1 Val/Val
variant to be associated with greater asthma prevalence and increased risk of
O3-associated respiratory morbidity (see discussion in Section 6.2.4.1).
The CHS also provided evidence of interactions between O3 exposure and variants in
genes for arginase (Salam et al., 2009). Arginase catalyzes the conversion of L-
arginine. Because L-arginine is a precursor of NO, higher arginase activity can limit
production of NO and subsequent nitrosative stress. Epidemiologic evidence of
associations of arginase variants with asthma are limited (Li et al., 2006a); however,
asthmatic subjects have been found to have higher arginase activity than non-
asthmatic subjects (Morris et al., 2004). The modifying effect of O3 and atopy on the
association between ARG1 and ARG2 haplotypes and asthma were evaluated using
likelihood ratio tests with appropriate interaction terms. Having more copies of the
ARGlh4 haplotype (compared to having zero copies) was associated with lower
odds of asthma, particularly among children with atopic asthma living in high O3
communities (OR: 0.12; [95% CI: 0.04, 0.43]). Having more copies of the ARG2h3
haplotype (compared to having zero copies) was associated with increased risk of
childhood-onset asthma among children in both low and high O3 communities.
The implications of findings are somewhat limited because the functional relevance
of the ARG1 and ARG2 variants is not clear.
7.2.1.2 Prevalence of Asthma and Asthma Symptoms
Some cross-sectional studies reviewed in the 2006 O3 AQCD observed positive
relationships between chronic exposure to O3 and prevalence of asthma and
asthmatic symptoms in school children (Ramadour et al., 2000; Wang et al., 1999)
while others (Kuo et al., 2002; Charpin et al., 1999) did not. Recent studies provide
additional evidence.
In a cross-sectional nationwide study of 32,672 Taiwanese school children, Hwang et
al. (2005) assessed the effects of air pollutants on the risk of asthma. The study
population was recruited from elementary and middle schools within 1 km of air
monitoring stations. The risk of asthma was related to O3 in the one-pollutant model.
The addition of other pollutants (NOX, CO, SO2, and PMi0), in two-pollutant and
7-8
-------
three-pollutant models, increased the O3 risk estimates. The prevalence of childhood
asthma was assessed in Portugal by contrasting the risk of asthma between a high O3
rural area and an area with low O3 levels (Sousaet al., 2011; Sousaet al., 2009;
Sousa et al., 2008). The locations were selected to provide a difference in O3 levels
without the confounding effects of other pollutants. Both evaluation for asthma
symptoms and FEVi suggested that O3 increased asthma prevalence. Clark et al.
(2010) investigated the effect of exposure to ambient air pollution in utero and during
the first year of life on risk of subsequent incidence asthma diagnosis up to 3-4 years
of age in a population-based nested case-control study for all children born in
southwestern British Columbia in 1999 and 2000 (n = 37,401; including 3,482
[9.3%] with asthma). Air pollution exposure for each subject was estimated based on
their residential address history using regulatory monitoring data, land use regression
modeling, and proximity to stationary pollutant sources. Daily values from the three
closest monitors within 50 km were used to calculate exposures. Traffic-related
pollutants were associated with the highest risk. Ozone was inversely correlated with
the primary traffic-related pollutants (r = -0.7 to -0.9). The low reliability of asthma
diagnosis in infants makes this study difficult to interpret (Martinez et al.. 1995). In a
cross-sectional analysis, Akinbami et al. (2010) examined the association between
chronic exposure to outdoor pollutants (12-month avg levels by county) and asthma
outcomes in a national sample of children ages 3-17 years living in U.S. metropolitan
areas (National Health Interview Survey, N = 34,073). A 5-ppb increase in estimated
8-h max O3 concentration (annual average) yielded a positive association for both
currently having asthma and for having at least 1 asthma attack in the previous year,
while the adjusted odds ratios for other pollutants were not statistically significant.
Models in which pollutant value ranges were divided into quartiles produced
comparable results. Multipollutant models (SO2 and PM) produced similar results.
The median value for 12-month avg O3 levels was 39.5 ppb and the IQR was
35.9-43.7 ppb. The adjusted odds for current asthma for the highest quartile
(49.9-59.5 ppb) of estimated O3 exposure was 1.56 (95% CI: 1.15, 2.10) with a
positive concentration-response relationship apparent from the lowest quartile to the
highest. Thus, this cross-sectional analysis and Hwang et al. (2005) provided further
evidence relating O3 exposure and the risk of asthma.
Relationships between long-term exposure and respiratory symptoms in asthmatic
children also were examined in the CHS. McConnell et al. (1999) examined the
association between O3 levels and the prevalence of chronic lower respiratory tract
symptoms in 3,676 cohort children with asthma. In this cross-sectional study,
bronchitis, phlegm, and cough were not associated with annual mean 1-h max O3
concentrations in children with asthma or wheeze. All other pollutants examined
(PMio, PM2.s, NO2, and gaseous acid) were associated with an increase in phlegm
but not cough. The mean annual average 1-h max O3 concentration was 65.6 ppb
(range 35.5 to 97.5) across the 12 communities. In another CHS analysis, McConnell
et al. (2003) evaluated relationships between air pollutants and bronchitic symptoms
among 475 children with asthma. The mean 4-year average 8-h avg O3 (10 a.m.-6
p.m.) concentration was 47.2 ppb (range 28.3 to 65.8) across the 12 communities.
For a 10-ppb increase in 8-h avg O3 averaged over 4 years, the between-community
odds ratio was 0.90 (95% CI: 0.82, 1.00) whereas the within-community
7-9
-------
(i.e., difference between one- and four-year average) odds ratio was larger, i.e., 1.79
(95% CI: 1.00, 3.21). The authors commented that if the larger within-community
effect estimates were correct, then other cross-sectional (between-community)
studies might have underestimated the true effect of air pollution on bronchitic
symptoms in children. These differences might be attributable to confounding by
poorly measured or unmeasured risk factors that vary between communities. Within
community effects may more accurately represent risk associated with pollutant
exposure because the analyses characterize health effects associated with changing
pollutant concentrations within a community, thereby minimizing potential
confounding by factors that are constant over time within a community. PM2.5, NO2,
and organic carbon also were associated with bronchitic symptoms. In two-pollutant
models, the within-community effect estimates for O3 were markedly reduced and no
longer statistically significant in some cases.
CHS also examined interactions between TNF-oc 308 genotype and long-term O3
exposure in the occurrence of bronchitic symptoms among children with asthma (Lee
et al.. 2009b). Increased airway levels of the cytokine TNF-oc has been related to
inflammation, and the GG genotype has been linked to lower expression of TNF-oc.
Asthmatic children with the GG genotype had a lower prevalence of bronchitic
symptoms compared with children carrying at least one A-allele (e.g., GA or AA
genotype). Low-versus high-O3 strata were defined as less than or greater than 50-
ppb O3 avg. Asthmatic children with TNF-308 GG genotype had a significantly
reduced risk of bronchitic symptoms with low-O3 exposure (OR: 0.53 [95% CI: 0.31,
0.91]). The risk was not reduced in children living in high-O3 communities (OR:
1.42 [95% CI: 0.75, 2.70]). The difference in genotypic effects between low- and
high-O3 environments was statistically significant among asthmatics (P for
interaction = 0.01), but not significant among non-asthmatic children. Figure 7-2
presents adjusted O3 community-specific regression coefficients plotted against
ambient O3 concentration, using weights proportional to the inverse variance.
Investigators further reported no substantial differences in the effect of the GG
genotype on bronchitic symptoms by long-term exposure to PMi0, PM2.5, NO2, acid
vapor, or second-hand smoke.
7-10
-------
Q.
f
!»
o §
0 -S.
co E
8 5T
LL. .0
OQ.
T3 '
s,
^
f
0
-2
20 30 40 50 60
Average ozone from 10 a.m. to 6 p.m. in communities (ppb)
70
Note: Using indicator variables for each category of genotype and O3 exposure, investigators calculated effect estimates for
TNF-a GG genotype on the occurrence of bronchitic symptoms among children with asthma.
Source: Reprinted with permission of John Wiley & Sons, (Lee et al., 2009b).
Figure 7-2 Ozone modifies the effect of TNF GG genotype on bronchitic
symptoms among children with asthma in the CHS.
Another CHS analyses reported interrelationships between variants in CAT and
myeloperoxidase (MPO) genes, ambient pollutants, and respiratory-related school
absences for 1,136 Hispanic and non-Hispanic white cohort children (Wenten et al.,
2009). A related study (Gilliland et al., 2001), found increased O3 exposure to be
related to greater school absenteeism due to respiratory illness but did not consider
genetic variants. Wenten et al. (2009) hypothesized that variation in the level or
function of antioxidant enzymes would modulate respiratory illness risk, especially
under high levels of oxidative stress expected from high ambient O3 exposure.
The joint effect of variants in these two genes (genetic epistasis) on respiratory
illness was examined because the enzyme products operate on the same substrate
within the same biological pathway. Risk of respiratory-related school absences was
elevated for children with CAT GG plus MPO GA or AA genotypes (RR: 1.35
[95% CI: 1.03, 1.77] compared to GG for both genes) and reduced for children with
CAT GA or AA plus MPO GA or AA (RR: 0.81 [95% CI: 0.55, 1.19] compared to
GG for both genes). Both CAT GG and MPO GA (or AA) genotypes produced a less
activity enzyme. In analyses that stratified communities into high and low O3
exposure groups by median levels (46.9 ppb), the protective effect of CAT GA or
AA plus MPO GA or AA genotype was largely limited to children living in
communities with high ambient O3 levels (RR: 0.42 [95% CI: 0.20, 0.89]).
The association of respiratory-illness absences with functional variants in CAT and
7-11
-------
MPO that differed by air pollution levels illustrates the need to consider genetic
epistasis in assessing gene-environment interactions.
Collective evidence from CHS provides an important demonstration of
gene-environment interactions. In the complex gene-environment setting a modifying
effect might not be reflected in an exposure main effect. The simultaneous
occurrence of main effect and interaction effect can occur. The study of
gene-environment interactions helps to dissect disease mechanisms in humans by
using information on susceptibility genes to focus on the biological pathways that are
most relevant to that disease (Hunter, 2005).
The French Epidemiology study on Genetics and Environment of Asthma (EGEA)
investigated the relationship between ambient air pollution and asthma severity in a
cohort in five French cities (Paris, Lyon, Marseille, Montpellier, and Grenoble)
(Rage et al.. 2009b). In this cross-sectional study, asthma severity over the past
12 months was assessed among 328 adult asthmatics using two methods: (1) a four-
class severity score that integrated clinical events and type of treatment; and (2) a
five-level asthma score based only on symptoms. Two measures of exposure were
also assessed: (1) [first method] closest monitor data from 1991 to 1995 where a total
of 93% of the subjects lived within 10 km of a monitor, but where 70% of the O3
concentrations were back-extrapolated values; and (2) [second method] a validated
spatial model that used geostatistical interpolations and then assigned air pollutants
to the geocoded residential addresses of all participants and individually assigned
exposure to ambient air pollution estimates. Higher asthma severity scores were
significantly related to both the 8-h avg O3 during April-September and the number
of days with 8-h O3 averages above 55 ppb. Both exposure assessment methods and
severity score methods resulted in very similar findings. Effect estimates of O3 were
similar in three-pollutant models. No PM data were available. Since these estimates
were not sensitive to the inclusion of ambient NO2 in the three-pollutant models, the
authors viewed the findings not to be explained by particles which usually have
substantial correlations between PM and NO2. Effect estimates for O3 in three-
pollutant models including O3, SO2, and NO2 yielded OR for O3-days of 2.74
(95% CI: 1.68, 4.48) per IQR days of 10-28 (+18) ppb. The effect estimates for SO2
and NO2 in the three-pollutant model were 1.33 (95% CI: 0.85, 2.11) and 0.94
(95% CI: 0.68, 1.29), respectively. Taking into account duration of residence did not
change the result. This study suggests that a higher asthma severity score is related to
long-term O3 exposure.
An EGEA follow-up study (Jacquemin et al.. 2012). examines the relationship
between asthma and O3, NO2, and PMi0. New aspects considered include: (1)
examination of three domains of asthma control (symptoms, exacerbations, and lung
function); (2) levels of asthma control (controlled, partially controlled, and
uncontrolled asthma); and (3) PMi0 and multipollutant analysis. In this cross-
sectional analysis, EGEA2 studied 481 adult subjects with current asthma from 2003
to 2007. The IQRs were 11 (41-52) ng/m3 for annual O3 and 13 (25-38) ng/m3 for
summer (April-September) O3. The association between asthma control and air
pollutants was expressed by ORs (reported for one IQR of the pollutant), derived
7-12
-------
from multinomial logistic regression. For each factor, the simultaneous assessment of
the risk for uncontrolled asthma and for partly controlled asthma was compared with
controlled asthma using a composite of the three domains. In crude and adjusted
models, O3-sum and PMi0 were positively associated with partly controlled and
uncontrolled asthma, with a clear gradient from controlled, partly controlled
(OR = 1.53, 95% CI: 1.01, 2.33) and uncontrolled (OR = 2.14, 95% CI: 1.34, 3.43)
(from the multinomial logistic regression).
Separately, they used a composite asthma control classification that used the ordinal
logistic regression for risk comparing controlled to partly controlled asthma and
comparing partly controlled to uncontrolled asthma. For these two pollutants, the
ORs assessed using the ordinal logistic regression were significant (ORs were 1.69
(95% CI: 1.22, 2.34) and 1.35 (95% CI: 1.13, 1.64) for O3-sum andPM10,
respectively). For two pollutant models using the ordinal logistic regression, the
adjusted ORs for O3-sum and PM10 included simultaneously in a unique model were
1.50 (95% CI: 1.07, 2.11) for O3-sum and 1.28 (95% CI: 1.06, 1.55) for PM10,
respectively. This result suggests that the effects of both pollutants are independent.
The analysis of the associations between air pollution for all asthma subjects and
each one of the three asthma control domains showed the following: (1) for lung
function defined dichotomously as percent predicted FEVi value =80
(OR = 1.35, 95% CI: 0.80, 2.28 for adjusted O3-sum); (2) for symptoms defined as
asthma attacks or dyspnea or woken by asthma attack or shortness of breath in the
past three months (OR = 1.59, 95% CI: 1.10, 2.30 for adjusted O3-sum); and (3) for
exacerbations defined at least one hospitalizations or ER visits in the last year or oral
corticosteroids in the past three months (OR = 1.58, 95% CI: 0.97, 2.59 for adjusted
O3-sum). Since the estimates for both pollutants were more stable and significant
when using the integrated measure of asthma control, this indicates that the results
are not driven by one domain. These results support an effect of long-term exposure
to O3 on asthma control in adulthood in subjects with pre-existing asthma.
Goss et al. (2004) investigated the effect of O3 on pulmonary exacerbations and lung
function in individuals over the age of 6 years with cystic fibrosis (n = 11,484).
The study included patients enrolled in the Cystic Fibrosis Foundation National
Patient Registry, which contains demographic and clinical data collected annually at
accredited centers for cystic fibrosis. For 1999 through 2000, the annual mean O3
concentration, calculated from 1-h averages from 616 monitors in the U.S. EPA
Aerometric Information Retrieval System (AIRS), was 51.0 ppb (SD 7.3). Exposure
was assessed by linking air pollution values from AIRS with the patient's home ZIP
code. No clear association was found between annual mean O3 and lung function
parameters. However, a 10 ppb increase in annual mean O3 was associated with a
10% (95% CI: 3, 17) increase in the odds of two or more pulmonary exacerbations.
Significant excess odds of pulmonary exacerbations also were observed with
increased annual mean PMi0 and PM2.5 concentrations. The O3 effect was robust to
adjustment for PMi0 and PM2.5, 8% (95% CI: 1, 15) increase in odds of two or more
pulmonary exacerbations per 10 ppb increase in annual mean O3.
7-13
-------
7.2.2 Asthma Hospital Admissions and ED Visits
The studies on O3-related hospital discharges and emergency department (ED) visits
for asthma and respiratory disease that were available in the 2006 O3 AQCD mainly
looked at the daily time metric. Collectively the short-term O3 studies presented
earlier in Section 6.2.7.5 indicate that there is evidence for increases in both hospital
admissions and ED visits related to all respiratory outcomes including asthma with
stronger associations in the warm months. New studies evaluated long-term O3
exposure metrics, providing a new line of evidence that suggests a positive exposure-
response relationship between first asthma hospital admission and long-term O3
exposure.
An ecologic study (Moore et al.. 2008) evaluated time trends in associations between
declining warm-season O3 concentrations and hospitalization for asthma in children
in California's South Coast Air Basin who ranged in age from birth to 19 years.
Quarterly average concentrations from 195 spatial grids, 10x10 km, were used.
Ozone was the only pollutant associated with increased hospital admissions over the
study period. A linear relation was observed for asthma hospital discharges (Moore
et al.. 2008). A matched case-control study (Karr et al.. 2007) was conducted on
infant bronchiolitis (ICD 9, code 466.1) hospitalization and two measures of long-
term pollutant exposure (the month prior to hospitalization and the lifetime average)
for O3 in the South Coast Air Basin of southern California among 18,595 infants
born between 1995 and 2000. Ozone was associated with reduced risk in the single-
pollutant model, but this relation did not persist in multipollutant models (CO, NO2,
andPM2.5).
In a cross-sectional study, Meng et al. (2010) examined associations between air
pollution and asthma morbidity in the San Joaquin Valley in California by using the
2001 California Health Interview Survey data from subjects ages 1 to 65+ who
reported physician-diagnosed asthma (n = 1,502). Subjects were assigned annual
average concentrations for O3 based on residential ZIP code and the closet air
monitoring station within 8 km but did not have data on duration of residence.
Multipollutant models for O3 and PM did not differ substantially from single-
pollutant estimates, indicating that pollutant multi-collinearity is not a problem in
these analyses. The authors reported increased asthma-related ED visits or
hospitalizations for O3 (OR = 1.49; [95% CI: 1.05, 2.11] per 10 ppb) for all ages.
Positive associations were obtained for symptoms, but 95% confidence intervals
included null values. Associations for symptoms for adults (ages 18+) were observed
(OR = 1.40; [95% CI: 1.02, 1.91] per 10 ppb).
Associations between air pollution and poorly controlled asthma among adults in
Los Angeles and San Diego Counties were investigated using the California Health
Interview Survey data collected between November 2000 and September 2001
(Meng et al., 2007). Each respondent was assigned an annual average concentration
measured at the nearest station within 5 miles of the residential cross-street
intersection. Poorly controlled asthma was defined as having daily or weekly asthma
symptoms or at least one ED visit or hospitalization because of asthma during the
7-14
-------
past 12 months. This cross-sectional study reports an OR of 3.34 (95% CI: 1.01,
11.09) for poorly controlled asthma when comparing those 65 years of age and older
above the 90th percentile (28.7 ppb) level to those below that level. Copollutant
(PM) analysis produced similar results.
Evidence associating long-term O3 exposure to first asthma hospital admission in a
concentration-response relationship is provided in a retrospective cohort study (Lin et
al.. 2008b). This study investigated the association between chronic exposure to O3
and childhood asthma admissions (defined as a principal diagnosis of ICD9, code
493) by following a birth cohort of 1,204,396 eligible births born in New York State
during 1995-1999 to first asthma admission or until 31 December 2000. There were
10,429 (0.87%) children admitted to the hospital for asthma between 1 and 6 years of
age. The asthma hospitalization rate in New York State in 1993 was 2.87 per 1,000
(Lin et al., 1999). Three annual indicators (all 8-h max from 10:00 a.m. to 6:00 p.m.)
were used to define chronic O3 exposure: (1) mean concentration during the follow-
up period (41.06 ppb); (2) mean concentration during the O3 season (50.62 ppb); and
(3) proportion of follow-up days with O3 levels >70 ppb. In this study the authors
aimed to predict the risk of having asthma admissions in a birth cohort, but the time
to the first admission in children that is usually analyzed in survival models was not
their primary interest. The effects of copollutants were assessed and controlled for
using the Air Quality Index (AQI). Interaction terms were used to assess potential
effect modifications. A positive association between chronic exposure to O3 and
childhood asthma hospital admissions was observed indicating that children exposed
to high O3 levels over time are more likely to develop asthma severe enough to be
admitted to the hospital. The various factors were examined and differences were
found for younger children (1-2 years), poor neighborhoods, Medicaid/self-paid
births, geographic region and others. As shown in Figure 7-3. positive concentration-
response relationships were observed. Asthma admissions were significantly
associated with increased O3 levels for all chronic exposure indicators (ORs,
1.16-1.68). When estimating the O3 effect using the exceedance proportion, an
increase was observed (OR = 1.68; [95% CI: 1.64, 1.73]) in hospital admissions with
an IQR (2.51%) increase in O3. A proportional hazards model for the New York City
data was run as a sensitivity analysis and it yielded similar results between asthma
admissions and chronic exposure to O3 (Cox model: HR = 1.14, [95% CI: 1.124,
1.155] is similar to logistic model results: OR =1.16 [95% CI: 1.15, 1.171) (Lin.
2010). Thus, this study provides evidence associating long-term O3 exposure to first
asthma hospital admission in a concentration-response relationship.
In considering relationships between long-term pollutant exposure and chronic
disease health endpoints, Kunzli (2012) offers two hypotheses relevant to research on
air pollution and chronic disease where chronic pathologies are found with acute
expressions of the chronic disease: "HI: Exposure provides a basis for the
development of the underlying chronic pathology, which increases the pool of people
with chronic conditions prone to exacerbations; H2: Exposure triggers an acute event
(or a state of frailty that results in an event with a delay of a few days or weeks)
among those with the disease." Kiinzli (2012) states if associations of pollution with
events are much larger in the long-term studies, it provides some indirect evidence in
7-15
-------
support of HI. If air pollution increases the pool of subjects with the chronic
pathology (HI), more acute events are expected to be seen for higher exposures since
events due to various causes are part of the chronic disease pathway.
Kunzli (2012) makes such a comparison noting larger associations with long-term
NO2 exposures for adult asthma hospital admissions (Andersen et al.. 2012) as
compared to short-term NO2 exposures for asthma hospital admissions (Peel et al.,
2005). In a further example, Pope (2007) makes similar conclusions comparing long-
term PM mortality study results to short-term PM mortality studies. The results of
Lin et al. (2008b) for first asthma hospital admission, presented below, show effect
estimates that are larger than those reported in a study of asthma hospital admissions
in New York State by Silverman and Ito (2010), discussed in Chapter 6 (both studies
are for young children). This provides some support for the hypothesis that O3
exposure may not only have triggered the events but also increased the pool of
asthmatics. However, caution is warranted in attributing associations in the Lin et al.
(2008b) study to long-term exposures since there is potential for short-term
exposures to contribute to the observed associations.
c in Low exposure
3.0
2.5
2.0
1.5
1.0
0.5
n
0-33%
CHD Medium exposure 34-66%
i i High exposure
^ 67%
2.06
(1.87-2.27)
1.69 1.64 T
1.43 (1.52-1
(1.J9-1.R8) -T
1.00
(ref)
y
.80) (1.48-1.82)
T
1.00
(ref)
1
-T
New York City
Other NYS regions
Regions
Note: Adjusted for child's sex, age, birth weight, and gestational age; maternal race, ethnicity, age, education, insurance, and
smoking status during pregnancy; and regional poverty level and temperature. ORs by low, medium, and high exposure are
shown for New York City (NYC: low [37.3 ppb], medium [37.3-38.11 ppb], high [38.11+ ppb] and other New York State regions
(Other NYS regions: low [42.58 ppb], medium [42.58-45.06 ppb], high [45.06+ ppb]) for first asthma hospital admission.
Source: Lin (2010): Lin et al. (2008b)
Figure 7-3 Ozone-asthma concentration-response relationship using the mean
concentration during the entire follow-up period for first asthma
hospital admission.
7-16
-------
7.2.3 Pulmonary Structure and Function
7.2.3.1 Pulmonary Structure and Function: Evidence from
Epidemiology Studies
The definitive 8-year follow-up analysis of the first cohort of the CHS, which is
discussed in Section 7.2 (Gauderman et al.. 2004). provided little evidence that long-
term exposure to ambient O3 was associated with significant deficits in the growth
rate of lung function in children. A later CHS study (Islam et al.. 2007) examined
relationships between air pollution, lung function, and new onset asthma and
reported no substantial differences in the effect of O3 on lung function. Ozone
concentrations from the least to most polluted communities (mean annual average of
8-h avg O3) ranged from 30 to 65 ppb, as compared to the ranges observed for the
other pollutants, which had 4-fold- to 8-fold differences in concentrations. In a more
recent CHS study, Breton et al. (2011) hypothesized that genetic variation in genes
on the glutathione metabolic pathway may influence the association between ambient
air pollutant exposures and lung function growth in children. They investigated
whether genetic variation in glutathione genes GSS, GSR, GCLC, and GCLM was
associated with lung function growth in healthy children using data collected on
2,106 children over an 8-year time-period as part of the Children's Health Study.
Breton et al. (2011) found that variation in the GSS locus was associated with
differences in risk of children for lung function growth deficits associated with NO2,
PMio, PM2.5, elemental carbon, organic carbon, and O3. The negative effects of air
pollutants were largely observed within participants who had a particular GSS
haplotype. The effects ranged from -124.2 to -149.1 mL for FEVi, -92.9 to
-126.7 mL for FVC and -193.9 to -277.9 mL/sec for MMEF for all pollutants except
O3, for which some positive associations were reported: 25.9 mL for FEVi; 0-1 mL
for FVC, and 166.5 mL/sec for MMEF. Ozone was associated with larger decreases
in lung function in children without this haplotype, when compared to the other
pollutants with values of-76.6 mL for FEVi, -17.2 mL for FVC, and -200.3 mL/sec
for MMEF, but only the association with MMEF was statistically significant.
As discussed in the 2006 O3 AQCD, a study of freshman students at the University
of California, Berkeley reported an interaction between lifetime exposure to O3 and
baseline FEF2s-75/FVC ratio, a measure of intrinsic airway size for decreased
measures of small airways (<2 mm) function (FEF75 and FEF25-75) (Tager et al..
2005). Subjects with a small ratio (indicating an increased airway size relative to
their lung volume) had decreases in FEF75 and FEF 25.75 for increases in lifetime
exposure to O3. Kinney and Lippmann (2000) examined 72 nonsmoking adults
(mean age 20 years) from the second-year class of students at the U.S. Military
Academy in West Point, NY, and reported results that appear to be consistent with a
decline in lung function that may in part be due to O3 exposures over a period of
several summer months. Ihorst et al. (2004) examined 2,153 children with a median
age of 7.6 years and reported pulmonary function results which indicated that
significantly lower FVC and FEVi increases were associated with higher O3
7-17
-------
exposures over the medium-term of several summer months, but not over several
months in the winter. Semi-annual mean O3 concentrations ranged from 22 to 54 ppb
during the summer months and 4 to 36 ppb during the winter months. Further, over
the longer-term 3.5-year period Ihorst et al. (2004) found that higher mean summer
months O3 levels were not associated with growth rates in lung function and for FVC
and FEVi, in contrast to the significant medium-term effects. Frischer et al. (1999)
found that higher O3 over one summer season, one winter season, and greater
increases from one summer to the next over a three-year period were associated with
smaller increases in lung function growth, indicating both medium and longer-term
effects.
Mortimer et al. (2008a, b) examined the association of prenatal and lifetime
exposures to air pollutants with pulmonary function and allergen sensitization in a
subset of asthmatic children (ages 6-11) included in the Fresno Asthmatic Children's
Environment Study (FACES). Monthly means of pollutant levels for the years
1989-2000 were created and averaged separately across several important
developmental time-periods, including: the entire pregnancy, each trimester, the first
3 years of life, the first 6 years of life, and the entire lifetime. In the first analysis
(Mortimer et al., 2008a), negative effects on pulmonary function were found for
exposure to PMi0, NO2, and CO during key neonatal and early life developmental
periods. The authors did not find a negative effect of exposure to O3 within this
cohort. In the second analysis (Mortimer et al., 2008b), sensitization to at least one
allergen was associated, in general, with higher levels of CO and PMi0 during the
entire pregnancy and second trimester, and higher PMi0 during the first 2 years of
life. Lower exposure to O3 during the entire pregnancy or second trimester was
associated with an increased risk of allergen sensitization. Although the pollutant
metrics across time periods were correlated, the strongest associations with the
outcomes were observed for prenatal exposures. Though it may be difficult to
disentangle the effect of prenatal and postnatal exposures, the models from this group
of studies suggest that each time period of exposure may contribute independently to
different dimensions of school-aged children's pulmonary function. For 4 of the 8
pulmonary-function measures (FVC, FEVi, PEF, FEF25-75), prenatal exposures were
more influential on pulmonary function than early-lifetime metrics, while, in
contrast, the ratio of measures (FEVi/FVC and FEF2s-75/FVC) were most influenced
by postnatal exposures. When lifetime metrics were considered alone, or in
combination with the prenatal metrics, the lifetime measures were not associated
with any of the outcomes. This suggests that the timing of the O3 exposure may be
more important than the overall dose, and prenatal exposures are not just markers for
lifetime or current exposures.
Latzin et al. (2009) examined whether prenatal exposure to air pollution was
associated with lung function changes in the newborn. Tidal breathing, lung volume,
ventilation inhomogeneity and eNO were measured in 241 unsedated, sleeping
neonates (age = 5 weeks). Consistent with the previous studies, no association was
found for prenatal exposure to O3 and lung function.
7-18
-------
In a cross-sectional study of adults, Qian et al. (2005) examined the association of
long-term exposure to O3 and PMi0 with pulmonary function from data of 10,240
middle-aged subjects who participated in the Atherosclerosis Risk in Communities
(ARIC) study in four U.S. communities. A surrogate for long-term O3 exposure from
daily data was determined at the individual level. Ozone was significantly and
negatively associated with measures of pulmonary function.
To determine the extent to which long-term exposure to outdoor air pollution
accelerates adult decline in lung function, Forbes et al. (2009b) studied the
association between chronic exposure to outdoor air pollution and lung function in
approximately 42,000 adults aged 16 and older who were representatively sampled
cross-sectionally from participants in the Health Survey for England (1995, 1996,
1997, and 2001). FEVi was not associated with O3 concentrations. In contrast to the
results for PMi0, NO2, and SO2; combining the results of all the survey years showed
that a 5-ppb difference in O3 was counter-intuitively associated with a higher FEVi
by 22 mL.
In a prospective cohort study consisting of school-age, non-asthmatic children in
Mexico City (n = 3,170) who were 8 years of age at the beginning of the study,
Rojas-Martinez et al. (2007) evaluated the association between long-term exposure to
O3, PMio and NO2 and lung function growth every 6 months from April 1996
through May 1999. Exposure data were provided by 10 air quality monitor stations
located within 2 km of each child's school. Over the study period, 8-h O3
concentrations ranged from 60 ppb (SD, ± 25) in the northeast area of Mexico City to
90 ppb (SD, ± 34) in the southwest, with an overall mean of 69.8 ppb.
In multipollutant models, an IQR increase in mean O3 concentration of 11.3 ppb was
associated with an annual deficit in FEVi of 12 mL in girls and 4 mL in boys.
Single-pollutant models showed an association between ambient pollutants (O3,
PMio, and NO2) and deficits in lung function growth. While the estimates from
copollutant models were not substantially different than single pollutant models,
independent effects for pollutants could not be estimated accurately because the
traffic-related pollutants were correlated. To reduce exposure misclassification,
microenvironmental and personal exposure assessments were conducted in a
randomly selected subsample of 60 children using passive O3 samplers. Personal O3
concentrations were correlated (p <0.05) with the measurements obtained from the
fixed-site air monitoring stations.
In the 2006 O3 AQCD, few studies had investigated the effect of chronic O3
exposure on pulmonary function. The strongest evidence was for medium-term
effects of extended O3 exposures over several summer months on lung function
(FEVi) in children, i.e., reduced lung function growth being associated with higher
ambient O3 levels. Longer-term studies (annual), investigating the association of
chronic O3 exposure on lung function (FEVi) such as the definitive 8-year follow-up
analysis of the first cohort (Gauderman et al.. 2004) provide little evidence that long-
term exposure to ambient O3 at current levels is associated with significant deficits in
the growth rate of lung function in children. Analyses indicated that there was no
evidence that either 8-h avg O3 (10 a.m. to 6 p.m.) or 24-h avg O3 was associated
7-19
-------
with any measure of lung function growth over a 4-year (age 10 to 14 years;
(Gauderman et al., 2000)) or 8-year (age 10 to 18 years; (Gauderman et al., 2004))
period. However, most of the other pollutants examined (including PM2.5, NO2, acid
vapor, and elemental carbon) were found to be significantly associated with reduced
growth in lung function. In addition, there was only about a 2- to 2.5-fold difference
in O3 concentrations from the least to most polluted communities (mean annual
average of 8-h avg O3 ranged from 30 to 65 ppb), versus the ranges observed for the
other pollutants (which had 4- to 8-fold differences in concentrations).
Short-term O3 exposure studies presented in Section 6.2.1.2 provide a cumulative
body of epidemiologic evidence that strongly supports associations between ambient
O3 exposure and decrements in lung function among children. For new studies of
long-term O3 exposure relationship to pulmonary function, one study, where O3 and
other pollutant levels were higher (90 ppb at high end of the range) than those in the
CHS, observes a relationship between O3 concentration and pulmonary function
declines in school-aged children. Two studies of adult cohorts provide mixed results
where long- term exposures were at the high end of the range with levels of 49.5 ppb
in one study and 27 ppb IQR in the other. Toxicological studies examining monkeys
have provided data for airway resistance in an asthma model but this is difficult to
compare to FEVi results. Thus there is little new evidence to build upon the very
limited studies of pulmonary function (FEVi) from the 2006 O3 AQCD.
7.2.3.2 Pulmonary Structure and Function: Evidence from
Toxicological Studies and Nonhuman Primate Asthma
Models
Long-term studies in animals allow for greater insight into the potential effects of
prolonged exposure to O3, that may not be easily measured in humans, such as
structural changes in the respiratory tract. As reviewed in the 1996 and 2006 O3
AQCDs and Chapter 5_ of this ISA, there are both qualitative and quantitative
uncertainties in the extrapolation of data generated by rodent toxicology studies to
the understanding of health effects in humans. Despite these uncertainties,
epidemiologic studies observing functional changes in humans can attain biological
plausibility, in conjunction with long-term toxicological studies, particularly O3-
inhalation studies performed in non-human primates whose respiratory system most
closely resembles that of the human. An important series of studies have used
nonhuman primates to examine the effect of O3 alone or in combination with an
inhaled allergen, house dust mite antigen, on morphology and lung function. These
animals exhibit the hallmarks of allergic asthma defined for humans, including: a
positive skin test for HDMA with elevated levels of IgE in serum and IgE-positive
cells within the tracheobronchial airway walls; impaired airflow which is reversible
by treatment with aerosolized albuterol; increased abundance of immune cells,
especially eosinophils, in airway exudates and bronchial lavage; and development of
nonspecific airway responsiveness (NHLBI, 2007). Hyde et al. (2006) compared
asthma models of rodents (mice) and the nonhuman primate model to responses in
7-20
-------
humans and concluded that the unique responses to inhaled allergen shown in the
rhesus monkeys make it the most appropriate animal model of human asthma. These
studies and others have demonstrated changes in pulmonary function and airway
morphology in adult and infant nonhuman primates repeatedly exposed to
environmentally relevant concentrations of O3 (Joad et al.. 2008; Carey et al.. 2007;
Plopper et al.. 2007; Fanucchi et al.. 2006; Joad et al.. 2006; Evans et al.. 2004;
Larson et al.. 2004; Tran et al.. 2004; Evans et al.. 2003; Schelegle et al.. 2003;
Fanucchi et al.. 2000; Hydeetal. 1989; Harkema et al.. 1987a; Harkema et al..
1987b; Fujinaka et al.. 1985). Many of the observations found in adult monkeys have
also been noted in infant rhesus monkeys, although a direct comparison of the degree
of effects between adult and infant monkeys has not been reported. The findings of
these nonhuman primate studies have also been observed in rodent studies discussed
at the end of this section and included in Table 7-1.
The initial observations in adult nonhuman primates have been expanded in a series
of experiments using infant rhesus monkeys repeatedly exposed to 0.5 ppm O3
starting at 1 month of age1 (Plopper et al.. 2007). The purpose of these studies,
designed by Plopper and colleagues, was to determine if a cyclic regimen of O3
inhalation would amplify the allergic responses and structural remodeling associated
with allergic sensitization and inhalation in the infant rhesus monkey. In terms of
pulmonary function changes, after several episodic exposures of infant monkeys to
O3, they observed a significant increase in the baseline airway resistance, which was
accompanied by a small increase in airway responsiveness to inhaled histamine
(Schelegle et al.. 2003). although neither measurement was statistically different
from filtered air control values. Exposure of animals to inhaled house dust mite
antigen alone also produced small but not statistically significant changes in baseline
airway resistance and airway responsiveness, whereas the combined exposure to both
(O3 + antigen) produced statistically significant and greater than additive changes in
both functional measurements. This nonhuman primate evidence of an O3-induced
change in airway resistance and responsiveness supports the biologic plausibility of
long-term exposure to O3 contributing to the effects of asthma in children.
To understand which conducting airways and inflammatory mechanisms are
involved in O3-induced airway hyperresponsiveness in the infant rhesus monkey, a
follow-up study examined airway responsiveness ex vivo in lung slices (Joad et al..
2006). Using video microscopy to morphometrically evaluate the response of bronchi
and respiratory bronchioles to methacholine, (a bronchoconstricting agent commonly
used to evaluate airway responsiveness in asthmatics), the investigators observed
differential effects for the two airway sizes. While episodic exposure to O3 alone
(0.5 ppm) had little effect on ex vivo airway responsiveness in bronchi and
respiratory bronchioles, exposure to dust mite antigen alone produced airway
hyperresponsiveness in the large bronchi, whereas O3 + antigen produced significant
increases in airway hyperresponsiveness only in the respiratory bronchioles. These
1 Schelegle et al. (2003) used a two-by-two block design. Twenty-four infant rhesus monkeys (30 days old) were exposed to 11
episodes (total of 6-months exposure period) of filtered air (FA), house dust mite allergen (HDMA), O3 (5 days each followed by
9 days of FA). Ozone was delivered for 8h/day at 0.5 ppm. Twelve of the monkeys (HDMA, and HDMA + O3 groups) were
sensitized to house dust mite allergen (HDMA, confirmed by skin testing). To evaluate the potential for recovery, the 5 months of
exposure were followed by another 6 months in FA until the monkeys were reevaluated at 12 months of age.
7-21
-------
results suggest that effect of O3 on airway responsiveness occurs predominantly in
the smaller bronchioles, where dosimetric models indicate the dose would be higher.
The functional changes in the conducting airways of infant rhesus monkeys exposed
to either O3 alone or O3 + antigen were accompanied by a number of cellular and
morphological changes, including a significant 4-fold increase in eosinophils, (a cell
type important in allergic asthma), in the bronchoalveolar lavage of infant monkeys
exposed to O3 alone. Thus, these studies demonstrate both functional and cellular
changes in the lung of infant monkeys after cyclic exposure to 0.5 ppm O3. This
concentration, provides relevant information to understanding the potentially
damaging effects of ambient O3 exposure on the respiratory tract of humans.
No concentration-response data, however, are available from these nonhuman
primate studies.
In addition to these functional and cellular changes, significant structural changes in
the respiratory tract have been observed in infant rhesus monkeys exposed to O3.
During normal respiratory tract development, conducting airways increase in
diameter and length in the infant rhesus monkey. Exposure to O3 alone (5 days of
0.5 ppm O3 at 8 h/day, followed by 9 days of filtered air exposures for 11 cycles),
however, markedly affected the growth pattern of distal conducting airways
(Fanucchi et al.. 2006). Whereas the first alveolar outpocketing occurred at airway
generation 13 or 14 in filtered air-control infant monkeys, the most proximal
alveolarized airways occurred at an average of 10 airway generations in O3-exposed
monkeys. Similarly, the diameter and length of the terminal and respiratory
bronchioles were significantly decreased in O3-exposed monkeys. Importantly, the
O3-induced structural pathway changes persisted after recovery in filtered air for
6 months after cessation of the O3 exposures. These structural effects were
accompanied by significant increases in mucus goblet cell mass, alterations in
smooth muscle orientation in the respiratory bronchioles, epithelial nerve fiber
distribution, and basement membrane zone morphometry. These latter effects are
noteworthy because of their potential contribution to airway obstruction and airway
hyperresponsiveness which are central features of asthma.
Because many cellular and biochemical factors are known to contribute to allergic
asthma, the effect of exposure to O3 alone or O3 + antigen on immune system
parameters was also examined in infant rhesus monkeys. Mast cells, which
contribute to asthma via the release of potent proteases, were elevated in animals
exposed to antigen alone but O3 alone had little effect on mast cell numbers and the
response of animals exposed to O3 + antigen was not different from that of animals
exposed to antigen alone; thus suggesting that mast cells played little role in the
interaction between O3 and antigen in this model of allergic asthma (Van Winkle et
al.. 2010). Increases in CD4+ and CD8+ lymphocytes were observed at 6 months of
age in the blood and bronchoalveolar lavage fluid of infant rhesus monkeys exposed
to O3 + antigen but not in monkeys exposed to either agent alone (Miller et al..
2009). Activated lymphocytes (i.e., CD25+ cells) were morphometrically evaluated
in the airway mucosa and significantly increased in infant monkeys exposed to
antigen alone or O3 + antigen. Although O3 alone had no effect on CD25+ cells, it
7-22
-------
did alter the anatomic distribution of CD25+ cells within the airways. Ozone had
only a small effect on these sets of immune cells and did not produce a strong
interaction with an inhaled allergen in this nonhuman primate model.
In addition to alterations in the immune system, nervous system interactions with
epithelial cells are thought to play a contributing role to airway hyperresponsiveness.
A critical aspect of postnatal lung development is the laying of nerve axons with
specific connections serving to maintain lung homeostasis. Aberrant innervation
patterns may underlie allergic airways disease pathology and long-term decrements
in airway function. As noted in the 2006 O3 AQCD, exposure of infant rhesus
monkeys altered the normal development of neural innervation in the epithelium of
the conducting airways (Larson et al., 2004). Significant mean reductions in nerve
fiber density were observed in the midlevel airways of animals exposed to O3 alone
(49% reduction), and O3 + antigen (55% reduction). Moreover, the morphology of
nerve bundles was altered. The persistence of these effects was examined after a
6-month recovery period, and although nerve distribution remained atypical, there
was a dramatic increase in airway nerve density (hyperinnervation) (Kajekar et al.,
2007). Thus, in addition to structural, immune, and inflammatory effects, exposure to
O3 produces alterations in airway innervation which may contribute to O3-induced
exacerbation of asthma. Evaluation of the pathobiology of airway remodeling in
growing lungs of neonates using an animal model where exposure to allergen
generates reactive airway disease with all the hallmarks of asthma in humans
illustrates that exposure to O3 and allergen early in life produces a large number of
disruptions of fundamental growth and differentiation processes.
A number of studies in both nonhuman primates and rodents demonstrate that O3
exposure can increase collagen synthesis and deposition, inducing fibrotic-like
changes in the lung (Lastetal., 1994; Chang etal., 1992; Moffatt et al., 1987; Reiser
et al., 1987; Last et al., 1984). Increased collagen content is often associated with
elevated abnormal cross links that appear to be irreversible (Reiser et al., 1987).
Generally changes in collagen content have been observed in rats exposed to 0.5 ppm
O3 or higher, although extracellular matrix thickening has been observed in the lungs
of rats exposed to an urban pattern of O3 with daily peaks of 0.25 ppm for 38 weeks
(Chang et al., 1992; Chang et al., 1991). A more recent study using an urban pattern
of exposure to 0.5 ppm O3 demonstrated that O3-induced collagen deposition in mice
is dependent on the activity of TGF-(3 (Katre et al., 2011). Sex differences have been
observed with respect to increased centriacinar collagen deposition and crosslinking,
which was observed in female but not male rats exposed to 0.5 and 1.0 ppm O3 for
20 months (Last et al., 1994). Few other long-term exposure morphological studies
have presented sex differences and most only evaluated males.
As described in the 1996 and 2006 O3 AQCDs, perhaps the largest chronic O3 study
was an NIEHS-NTP/HEI funded rodent study conducted by multiple investigators
studying a number of different respiratory tract endpoints (Catalano et al.. 1995b).
Rats were exposed to 0.12, 0.5, or 1.0 ppm O3 for 6 h/day and 5 d/week for 20
months. The most prominent changes were observed in the nasal cavity where a large
fraction of O3 is absorbed. Alterations in nasal function (increased mucous flow) and
7-23
-------
structure (goblet cell metaplasia) were observed at 0.5 and 1.0 ppm but not 0.12 ppm
O3. In the lung, the centriacinar region (CAR) was the anatomical site most affected
by O3. The epithelial cell lining was changed to resemble that seen in respiratory
bronchioles and the interstitial volume was increased. Biochemical analyses
demonstrated increased collagen and glycoaminoglycans, an observation that
supported the structural changes. As in the nose, these changes were observed only at
the two highest exposure concentrations. Importantly, despite these morphologic and
biochemical changes after 20 months of exposure, detailed pulmonary function
testing revealed little to no measurable change in function. Thus, minor respiratory
tract changes were observed after chronic exposure to O3 up to 1.0 ppm in the F344
rat model.
It is unclear what the long-term impact of O3-induced structural changes may be.
Simulated seasonal (episodic) exposure studies suggest that such exposures might
have cumulative impacts, and a number of studies indicate that structural changes in
the respiratory system are persistent or irreversible. For example, O3-induced
hyperplasia was still evident in the nasal epithelia of rats 13 weeks after recovery
from 0.5 ppm O3 exposure (Harkema et al., 1999). In a study of episodic exposure to
0.25 ppm O3, Chang et al. (1992) observed no reversal of basement membrane
thickening in rat lungs up to 17 weeks post-exposure. Thickening of the sub-
basement membrane is one of the persistent structural features observed in human
asthmatics (NHLBI, 2007). Episodic exposure (0.25 ppm O3, every other month) of
young monkeys induced equivalent morphological changes compared to
continuously exposed animals, even though they were exposed for half the time and
evaluation occurred a month after exposure ceased as opposed to immediately (Tyler
et al., 1988). Notably, episodic O3 exposure increased total lung collagen content,
chest wall compliance, and inspiratory capacity, suggesting a delay in lung
maturation in episodically-exposed animals. These changes were in contrast to the
continuously exposed group, which did not differ from the air exposed group in these
particular parameters but did exhibit greater bronchiolitis than the episodically
exposed animals. In a study by Harkema and colleagues (Harkema et al.. 1993.
1987b). monkeys (both males and females) were acutely exposed for 8 h/day to
0.15 ppm O3 (6 days) or chronically to 0.15 ppm or 0.3 ppm O3 (90 days). For most
endpoints in the nasal cavity, the observed morphologic changes and inflammation
were greater in the monkeys exposed for 6 days compared to 90 days, whereas in the
respiratory bronchioles of the same animals, there were no significant time or
concentration dependent differences (increased epithelial thickness and proportion of
cuboidal cells) between the 6 and 90 day exposure groups.
Stokinger (1962) reported that chronic bronchitis, bronchiolitis, and emphysematous
and fibrotic changes develop in the lung tissues of mice, rats, hamsters, and guinea
pigs exposed 6 h/day, 5 days/week for 14.5 months to a concentration slightly above
1 ppm O3. Rats continuously exposed for 3 to 5 months to 0.8 ppm O3 develop a
disease that resembles emphysema, and they finally die of respiratory failure
(Stephens et al., 1976). Ozone results in a greater response of fibroblasts in the
lesion, thickening of the alveolar septae, and an increase in number of alveolar
macrophages in the proximal alveoli.
7-24
-------
Table 7-1 Respiratory effects in nonhuman primates and rodents resulting
from long-term O$ exposure.
Study
Model
(ppm)
Exposure Duration Effects
Pinkerton et al.
(1998): Harkema et
al.(1997a):
Harkema et al.
(1997b): Catalano et
al. (1995b):
Catalano et al.
(1995a): Chang et
al. (1995): Pinkerton
etal. (1995):
Stockstill et al.
(1995): Harkema et
al. (1994): Last etal.
(1994): Plopperet
al. (1994)
Herbert etal. (1996)
Rat, male and 0.12
female, 0 5
Fischer F344,
6-8 weeks old 1 -°
Mice, male and 0.12
female, B6C3F1 , 6-7 n ^n
i i i U.OU
weeks old,
6 h/day, 5 days/week Effects similar to (or a model of) early
for 20 months fibrotic human disease were greater
in the periacinar region than in
terminal bronchioles. Thickened
alveolar septa observed at 0.12 ppm
O3. Other effects (e.g., mucous cell
metaplasia in the nose, mild fibrotic
response in the parenchyma, and
increased collagen in CAR of
females) observed at 0.5 to 1 .0 ppm.
Some morphometric changes
(epithelial thickening and
bronchiolarization) occurred after 2 or
3 months of exposure to 1 .0 ppm.
6 h/day, 5 days/week Similar to the response of rats in the
for 24 and 30 months same study (see rat above). Effects
were seen in the nose and
centriacinar region of the lung at 0.5
and 1.0 ppm.
Chang etal. (1991)
Rat, male, F344,
6 weeks old
Continuous:
0.12 or 0.25
Episodic/urban:
baseline 0.06;
peak 0.25
Continuous: 12 h/day
for 6 weeks
Simulated urban
pattern; slow rise to
peak 9 h/day,
5 days/week,
13 weeks
Increased Type 1 and 2 epithelial
volume assessed by TEM. Linear
relationship observed between
increases in Type 1 epithelial cell
volume and concentration x time
product. Degree of injury not related
to pattern of exposure (continuous or
episodic).
Chang etal. (1992)
Rat, male, F344,
6 weeks old
baseline 0.06;
peak 0.25
Slow rise to peak
9 h/day, 5 days/week,
13 and 78 weeks
Recovery in filtered
air for 6 or 17 weeks
Progressive epithelial hyperplasia,
fibroblast proliferation, and interstitial
matrix accumulation observed using
TEM. Interstitial matrix thickening due
to deposition of basement membrane
and collagen fibers. Partial recovery
of interstitial matrix during follow-up
periods in air; but no resolution of
basement membrane thickening.
Barry etal. (1985.
1983)
Rat, male, 1 day old
or6 weeks old
0.12 (adults
only)
0.25
12 h/day for 6 weeks
Lung and alveolar development not
significantly affected. Increased Type
1 and 2 epithelial cells and AM in
CAR alveoli, thickened Type 1 cells
with smaller volume and less surface
coverage as assessed by TEM
(adults and juveniles). In adults,
smaller but statistically significant
similar changes at 0.12 ppm,
suggesting linear concentration-
response relationship. No statistically
significant age-related effects
observed.
Tyler etal. (1988)
Monkey; male,
Macaca fascicularis,
7 mo old
0.25
8 h/day, 7 days/week,
Daily for 18 mo or
episodically every
other month for 18 mo
Episodic group
evaluated 1 mo
postexposure
Increased collagen content, chest
wall compliance, and inspiratory
capacity in episodic group only.
Respiratory bronchiolitis in both
groups. Episodically exposed group
incurred greater alterations in
physiology and biochemistry and
equivalent changes in morphometry
even though exposed for half the time
as the daily exposure group.
Harkema et al.
Rat, male, Fischer
0.25
8 h/day, 7 days/week Mucous cell hyperplasia in nasal
7-25
-------
Study
Model
(ppm)
Exposure Duration Effects
F344/N HSD, 10-14
weeks old
0.5
for 13 weeks
epithelium after exposure to 0.25 and
0.5 ppm O3; still evident after 13
weeks recovery from 0.5 ppm O3
exposure.
Van Bree et al.
(2002)
Rat, male, Wistar, 7
weeks old,
n = 5/group
0.4
23.5 h/day for 1,3,7,
28,or 56 days
Acute inflammatory response in
BALF reached a maximum at day 1
and resolved within 6 days during
exposure. Centriacinar region
inflammatory responses throughout
O3 exposure with increased collagen
and bronchiolization still present after
a recovery period.
Katreetal. (2011)
Mice; male,
C57BL/6, 6-8 weeks
old
0.5
8 h/day, [5 days/week
O3, and 2 days filtered
air] for 5 or 10 cycles
Sustained elevation in TGF-p and
PAI-1 in lung (5 or 10 cycles);
elevated a-SMA and increased
collagen deposition in airway walls
(after 10 cycles). Collagen increase
shown to depend on TGF-(3.
Schelegle et al.
(2003):
Harkema et al.
(1993. 1987b)
Monkey; Rhesus, 0.5
30 days old3
Monkey; Macaca 0.15
radiata, M, F 03
2-6 years old
8 h/day for 5 days, Goblet cell metaplasia, increased
every 5 days for a total AHR, and increased markers of
of 1 1 episodes allergic asthma (e.g., eosinophilia)
were observed, suggesting that
episodic exposure to O3 alters
postnatal morphogenesis and
epithelial differentiation and
enhances the allergic effects of
house dust mite allergen in the lungs
of infant primates.
8 h/day for 90 days Significant increase in epithelial
thickness in respiratory bronchioles
which was accompanied by increase
in cuboidal cells; nasal lesions
consisted of ciliated cell necrosis and
secretory cell hyperplasia; no
concentration response effects
Larson et al. (2004)
Monkey; Macaca
mulatta, 30 days old3
0.5
Plopperet al. (2007) Monkey; Rhesus, 0.5
30 days old3
Fanucchi et al.
(2006)
Monkey; male 0.5
Rhesus,30 days old
Reiser etal. (1987) Monkey; male and 0.61
female Cynomolgus
6-7 mo old
3sex not reported
11 episodes of 5 days
each, 8 h/day followed
by 9 days of recovery
5 months of episodic
exposure; 5 days O3
followed by 9 days of
filtered air, 8h/day.
5 months of episodic
exposure; 5 days O3
followed by 9 days of
filtered air, 8h/day.
8 h/day for 1 year
O3 or O3 + house dust mite antigen
caused changes in density and
number of airway epithelial nerves in
small conducting airways. Suggests
episodic O3 alters pattern of neural
innervation in epithelial compartment
of developing lungs.
Non-significant increases airway
resistance and airway
responsiveness with O3 or inhaled
allergen alone. Allergen + O3
produced additive changes in both
measures.
Cellular changes and significant
structural changes in the distal
respiratory tract in infant rhesus
monkeys exposed to O3 postnatally.
Increased lung collagen content
associated with elevated abnormal
cross links that were irreversibly
deposited.
Collectively, evidence from animal studies strongly suggests that chronic O3
exposure is capable of damaging the distal airways and proximal alveoli, resulting in
lung tissue remodeling and leading to apparent irreversible changes. Potentially,
persistent inflammation and interstitial remodeling play an important role in the
7-26
-------
progression and development of chronic lung disease. Further discussion of the
modes of action that lead to O3-induced morphological changes can be found in
Section 5.3.7. The findings reported in chronic exposure animal studies offer insight
into potential biological mechanisms for the suggested association between seasonal
O3 exposure and reduced lung function development in children as observed in
epidemiologic studies (see Section 7.2.3). Discussion of mechanisms involved in
lifestage susceptibility and developmental effects can be found in Section 5.4.2.4.
7.2.4 Pulmonary Inflammation, Injury, and Oxidative Stress
The 2006 O3 AQCD stated that the extensive human clinical and animal
toxicological evidence, together with the limited epidemiologic evidence available,
suggests a causal role for O3 in inflammatory responses in the airways. Short-term
exposure epidemiologic studies discussed earlier in Section 6.2.3.2 show consistent
associations of O3 exposure and increased airway inflammation and oxidative stress.
Further discussion of the mechanisms underlying inflammation and oxi dative stress
responses can be found in Section 5.3.3. Though the majority of recent studies focus
on short-term exposures, several epidemiologic and toxicology studies of long-term
exposure add to observations of O3-induced inflammation and injury.
Inflammatory markers and peak expiratory pulmonary function were examined in 37
allergic children with physician-diagnosed mild persistent asthma in a highly
polluted urban area in Italy and then again 7 days after relocation to a rural location
with significantly lower pollutant levels (Renzetti et al.. 2009). The authors observed
a 4-fold decrease in nasal eosinophils and a statistically significant decrease in
fractional exhaled nitric oxide along with an improvement in lower airway function.
Several pollutants were examined, including PMi0, NO2, and O3, though pollutant-
specific results were not presented. These results are consistent with studies showing
that traffic-related exposures are associated with increased airway inflammation and
reduced lung function in children with asthma and contribute to the notion that this
negative influence may be rapidly reversible. Exhaled NO (eNO) has been shown to
be a useful biomarker for airway inflammation in large population-based studies
(Linn et al., 2009). Thus, while the time scale of 7 days between examinations for
eNO needs to be evaluated for appropriateness, the results suggest that inflammatory
responses are reduced when O3 levels are decreased.
Chest radiographs (CXR) of 249 children in Mexico City who were chronically
exposed to O3 and PM2.5 were analyzed by Calderon-Garciduenas et al. (2006). They
reported an association between chronic exposures to O3 and other pollutants and a
significant increase in abnormal CXR's and lung CTs suggestive of a bronchiolar,
peribronchiolar, and/or alveolar duct inflammatory process, in clinically healthy
children with no risk factors for lung disease. These CXR and CT results should be
viewed with caution because it is difficult to attribute effects to O3 exposure.
In a cross-sectional study, Wood et al. (2009) examined the association of outdoor air
pollution with respiratory phenotype (PiZZ type) in alpha 1-antitrypsin deficiency
7-27
-------
(a-ATD) from the U.K. a-ATD registry. This deficiency leads to exacerbated
responses to inflammatory stimuli. In total, 304 PiZZ subjects underwent full lung
function testing and quantitative high-resolution computed tomography to identify
the presence and severity of COPD - emphysema. Mean annual air pollution data for
2006 was matched to the location of patients' houses and used in regression models
to identify phenotypic associations with pollution controlling for covariates. Relative
trends in O3 levels were assessed to validate use of a single year's data to indicate
long-term exposure and validation; data showed good correlations between modeled
and measured data (Stedman and Kent 2008). Regression models showed that
estimated higher exposure to O3 exposure was associated with worse gas transfer and
more severe emphysema, albeit accounting for only a small proportion of the lung
function variability. This suggests that a gene-specific group demonstrates a long-
term O3 exposure effect.
The similarities of nonhuman primates to humans make them attractive models in
which to study the effects of O3 on the respiratory tract. The nasal mucous
membranes, which protect the more distal regions of the respiratory tract, are
susceptible to injury from O3. Carey et al. (2007) conducted a study of O3 exposure
in infant rhesus macaques, whose nasal airways closely resemble that of humans.
Monkeys were exposed either acutely for 5 days (8 h/day) to 0.5 ppm O3, or
episodically for several biweekly cycles alternating 5 days of 0.5 ppm O3 with 9 days
of filtered air (0 ppm O3), designed to mimic human exposure (70 days total). All
monkeys acutely exposed to O3 had moderate to marked necrotizing rhinitis, with
focal regions of epitheliar exfoliation, numerous infiltrating neutrophils, and some
eosinophils. The distribution, character, and severity of lesions in episodically
exposed monkeys were similar to that of acutely exposed animals. Neither group
exhibited the mucous cell metaplasia proximal to the lesions, observed in adult
monkeys exposed continuously to 0.3 ppm O3 in another study (Harkema et al..
1987a). Adult monkeys also exhibited attenuation of inflammatory responses with
continued daily exposure (Harkema et al.. 1987a). but inflammation did not resolve
over time in young episodically exposed monkeys (Carey et al.. 2011). Inflammation
in conducting airways has also been observed in rats chronically exposed to O3.
Using an agar-based technique to fill the alveoli so that only the rat bronchi are
lavaged, a 90-day exposure of rats to 0.8 ppm O3 (8 h/day) elicited significantly
elevated pro-inflammatory eicosanoids PGE2 and 12-HETE in the conducting airway
compared to filtered air-exposed rats (Schmelzer et al.. 2006).
Persistent inflammation and injury leading to interstitial remodeling may play an
important role in the progression and development of chronic lung disease. Chronic
airway inflammation is an important component of both asthma and COPD.
The epidemiological evidence supporting an association between long-term exposure
to O3 and inflammation or injury is limited. However, animal studies clearly
demonstrate O3-induced inflammation and injury, which may or may not attenuate
with chronic exposure depending on the model. Further discussion of how O3
initiates inflammation can be found in Section 5.3.3.
7-28
-------
7.2.5 Allergic Responses
The association of air pollutants with childhood respiratory allergies was examined
in the U.S. using the 1999-2005 National Health Interview Survey of approximately
70,000 children, and ambient air pollution data from the U.S. EPA, with monitors
within 20 miles of each child's residential block (Parker et al., 2009). The authors
examined the associations between the reporting of respiratory allergy or hay fever
and medium-term exposure to O3 over several summer months, controlling for
demographic and geographic factors. Increased respiratory allergy/hay fever was
associated with increased O3 levels (adjusted OR per 10 ppb = 1.20; [95% CI: 1.15,
1.26]). These associations persisted after stratification by urban-rural status, inclusion
of multiple pollutants (O3, SO2, NO2, PM), and definition of exposure by differing
exposure radii; smaller samples within 5 miles of monitors were remarkably similar
to the primary results. No associations between the other pollutants and the reporting
of respiratory allergy/hay fever were apparent. Ramadour et al. (2000) reported no
relationship between O3 levels and rhinitis symptoms and hay fever. Hwang et al.
(2006) report the prevalence of allergic rhinitis (adjusted OR per 10 ppb = 1.05;
[95% CI: 0.98, 1.12]) in a large cross-sectional study in Taiwan. In a large cross-
sectional study in France, Penard-Morand et al. (2005) reported a positive
relationship between lifetime allergic rhinitis and O3 exposure in a two-pollutant
model with NO2. These studies related positive outcomes of allergic response and O3
exposure but with variable strength for the effect estimates. A toxicological study
reported that five weeks of continuous exposure to 0.4 ppm O3 (but not 0.1 or
0.2 ppm O3) augmented sneezing and nasal secretions in a guinea pig model of nasal
allergy (lijima and Kobavashi. 2004). Nasal eosinophils, which participate in allergic
disease and inflammation, and allergic antibody levels in serum were also elevated
by exposure to concentrations as low as 0.2 ppm (lijima and Kobavashi. 2004).
Nasal eosinophils were observed to decrease by 4-fold in 37 atopic, mildly asthmatic
children 7 days after relocation from a highly polluted urban area in Italy to a rural
location with significantly lower pollutant levels (Renzetti et al., 2009).
Inflammatory and allergic effects of O3 exposure (30 day mean) such as increased
eosinophil levels were observed in children in an Austrian study (Frischer et al.,
2001). Episodic exposure of infant rhesus monkeys to 0.5 ppm O3 for 5 months
appears to significantly increase the number and proportion of eosinophils in the
blood and airways (lavage) [protocol described above in Section 7.2.3.2 for Fanucchi
et al. (2006)1 (Maniar-Hew et al., 2011). These changes were not evident at 1 year of
age (6 months after O3 exposure ceased). Increased eosinophils levels have also been
observed after acute or prolonged exposures to O3 in adult bonnet and rhesus
monkeys (Hyde et al.. 1992: Eustis et al.. 1981).
Total IgE levels were related to air pollution levels in 369 adult asthmatics in five
French centers using generalized estimated equations (GEE) as part of the EGEA
study described earlier (Rage et al.. 2009a). Geostatistical models were performed on
4x4 km grids to assess individual outdoor air pollution exposure that was assigned to
subject's home address. Ozone concentrations were positively related to total IgE
levels and an increase of 5 ppb of O3 resulted in an increase of 20.4% (95% CI: 3.0,
7-29
-------
40.7) in total IgE levels. Nearly 75% of the subjects were atopic. In two-pollutant
models including O3 and NO2, the O3 effect estimate was decreased by 25% while
the NO2 effect estimate was decreased by 57%. Associations were not sensitive to
adjustment for covariates or the season of IgE measurements. These cross-sectional
results suggest that exposure to O3 may increase total IgE in adult asthmatics.
Although very few toxicological studies of long-term exposure examining allergy are
available, short-term exposure studies in rodents and nonhuman primates
demonstrate allergic skewing of immune responses and enhanced IgE production.
Due to the persistent nature of these responses, the short-term toxicological evidence
lends biological plausibility to the limited epidemiologic findings of an association
between long-term O3 exposure and allergic outcomes.
7.2.6 Host Defense
Short-term exposures to O3 have been shown to cause decreases in host defenses
against infectious lung disease in animal models. Acute O3-induced suppression of
alveolar phagocytosis and immune functions observed in animals appears to be
transient and attenuated with continuous or repeated exposures, although chronic
exposure (weeks, months) has been shown to slow alveolar clearance. In an
important study investigating the effects of longer term O3 exposure on
alveolobronchiolar clearance, rats were exposed to an urban pattern of O3
(continuous 0.06 ppm, 7 days/week with a slow rise to a peak of 0.25 ppm and
subsequent decrease to 0.06 ppm over a 9 h period for 5 days/week) for 6 weeks and
were exposed 3 days later to chrysotile asbestos, which can cause pulmonary fibrosis
and neoplasia (Pinkerton et al.. 1989). After 30 days, the lungs of the O3-exposed
animals had twice the number and mass of asbestos fibers as the air-exposed rats.
However, chronic exposures of 0.1 ppm do not cause greater effects on infectivity
than short exposures, due to defense parameters becoming reestablished with
prolonged exposures. No detrimental effects were seen with a 120-day exposure to
0.5 ppm O3 on acute lung injury from influenza virus administered immediately
before O3 exposure started. However, O3 was shown to increase the severity of
postinfluenzal alveolitis and lung parenchymal changes (Jakab and Bassett, 1990).
A recent study by Maniar-Hew et al. (2011) demonstrated that the immune system of
infant rhesus monkeys episodically exposed to 0.5 ppm O3 for 5 months1 appeared to
be altered in ways that could diminish host defenses. Reduced numbers of circulating
leukocytes were observed, particularly polymorphonuclear leukocytes (PMNs) and
lymphocytes, which were decreased in the blood and airways (bronchoalveolar
lavage). These changes did not persist at 1 year of age (6 months postexposure);
rather, increased numbers of monocytes were observed at that time point. Challenge
with EPS, a bacterial ligand that activates monocytes and other innate immune cells,
elicited lower responses in O3-exposed animals even though the relevant reactive cell
population was increased. This was observed in both an in vivo inhalation challenge
1 Exposure protocol is described above in Section 7.2.3.2 for Fanucchi et al. (2006).
7-30
-------
and an ex vivo challenge of peripheral blood mononuclear cells. Thus a decreased
ability to respond to pathogenic signals was observed six months after O3 exposure
ceased, in both the lungs and periphery.
7.2.7 Respiratory Mortality
A limited number of epidemiologic studies have assessed the relationship between
long-term exposure to O3 and mortality. The 2006 O3 AQCD concluded that an
insufficient amount of evidence existed "to suggest a causal relationship between
chronic O3 exposure and increased risk for mortality in humans" (U.S. EPA, 2006b).
Though total and cardio-pulmonary mortality were considered in these studies,
respiratory mortality was not specifically considered. In the most recent follow-up
analysis of the ACS cohort (Jerrett et al., 2009), cardiopulmonary deaths were
subdivided into respiratory and cardiovascular, separately, as opposed to combined in
the Pope et al. (2002) work. A 10-ppb increment in exposure to O3 elevated the risk
of death from respiratory causes and this effect was robust to the inclusion of PM2.s.
The association between increased O3 concentrations and increased risk of death
from respiratory causes was insensitive to the use of a random-effects survival model
allowing for spatial clustering within the metropolitan area and state of residence,
and to adjustment for several ecologic variables considered individually.
Additionally, a recent study (Zanobetti and Schwartz, 2011) observed an association
between long-term exposure to O3 and elevated risk of mortality among Medicare
enrollees that had previously experienced an emergency hospital admission due to
COPD.
7.2.8 Summary and Causal Determination
The epidemiologic studies reviewed in the 2006 O3 AQCD detected no associations
between long-term (annual) O3 exposures and asthma-related symptoms, asthma
prevalence, or allergy to common aeroallergens among children after controlling for
covariates. Little evidence was available to relate long-term exposure to ambient O3
concentrations with deficits in the growth rate of lung function in children.
Additionally, limited evidence was available evaluating the relationship between
long-term O3 concentrations and pulmonary inflammation and other endpoints. From
toxicological studies, it appeared that O3-induced inflammation tapered off during
long-term exposures, but that hyperplastic and fibrotic changes remained elevated
and in some cases even worsened after a postexposure period in clean air. Episodic
exposures were also known to cause more severe pulmonary morphologic changes
than continuous exposure (U.S. EPA, 2006b).
The recent epidemiologic evidence base consists of studies using a variety of designs
and analysis methods evaluating the relationship between long-term exposure to
ambient O3 concentrations and measures of respiratory health effects and mortality
conducted by different research groups in different locations. See Table 7-2 for O3
7-31
-------
concentrations associated with selected studies. Table 7-2 is organized by
longitudinal and cross-sectional studies both presented alphabetically. The positive
results from various designs and locations support a relationship between long-term
exposure to ambient O3 concentrations and respiratory health effects and mortality.
Earlier studies reported associations of new-onset asthma and O3 in an adult cohort
in California (McDonnell et al., 1999a; Greer et al., 1993) but only in males. In the
CHS cohort of children in 12 Southern California communities, long-term exposure
to O3 concentrations was not associated with increased risk of developing asthma
(McConnell et al., 2010); however, greater outdoor exercise was associated with
development of asthma in children living in communities with higher ambient O3
concentrations (McConnell et al., 2002). Recent CHS studies examined interactions
among genetic variants, long-term O3 exposure, and new onset asthma in children.
These prospective cohort studies are methodologically rigorous epidemiology
studies, and evidence indicates gene-O3 interactions. These studies have provided
data supporting decreased risk of certain different genetic variants on new onset
asthma (e.g., HMOX-1, ARG) that is limited to children either in low (Islam et al.,
2008) or high (Salam et al., 2009) O3 communities. Gene-environment interaction
also was demonstrated with findings that greater outdoor exercise increased risk of
asthma in GSTP1 He/Tie children living in high O3 communities (Islam et al., 2009).
Biological plausibility for these these gene-O3 environment interactions is provided
by evidence that these enzymes have antioxidant and/or anti-inflammatory activity
and participate in well recognized modes of action in asthma pathogenesis. As O3 is
a source of oxidants in the airways, oxidative stress serves as the link among O3
exposure, enzyme activity, and asthma.
Studies using a cross-sectional design provide support for a relationship between
long-term O3 exposure and health effects in asthmatics. A long-term O3 exposure
study relates bronchitic symptoms to TNF-308 genotype asthmatic children with
ambient O3 exposure in the CHS (Lee et al., 2009b). A study in five French cities
reports effects on asthma severity related to long-term O3 exposure (Rage et al.,
2009b). A follow-up study of this cohort (Jacquemin et al., 2012) supports an effect
of cumulative long-term O3 exposure on asthma control in adulthood in subjects with
pre-existing asthma. Akinbami et al. (2010) and Hwang et al. (2005) provide further
evidence relating O3 exposures and the risk of asthma. For the respiratory health of a
cohort based on the general U.S. population, risk of respiratory-related school
absences was elevated for children with the CAT and MPO variant genes related to
communities with high ambient O3 levels (Wenten et al., 2009).
7-32
-------
Table 7-2 Summary of selected key new studies examining annual O3
exposure and respiratory health effects.
Study; Os Range
Health Effect; (ppb)
Location Annual Mean Os Concentration (ppb) Percentiles
Longitudinal
Islam et al. (2008):
New-onset asthma;
CHS
55.2 high vs. 38.4 low communities
10:00 a.m. to 6:00 p.m. average
See left
Islam et al. (2009):
New-onset asthma;
CHS
55.2 high vs. 38.4 low communities
10:00 a.m. to 6:00 p.m.
See left
Lin et al. (2008b):
First asthma hospital admission;
New York State -10 regions
Range of mean O3 concentrations over the
10 New York Regions 37.51 to 47.78
8-h max 10:00 a.m. to 6:00 p.m.
See left
Salam et al. (2009):
Childhood onset asthma;
CHS
Lee et al. (2009b):
Bronchitic symptoms in asthmatic children;
CHS
O3 greater than or less than 50 ppb
Above and below 50 ppb
See left
Cross-sectional
Akinbamietal. (2010):
Current asthma
U.S.
Hwang et al. (2005):
Prevalence of asthma;
Taiwan
Jacquemin et al. (2012):
Asthma control in adults;
Five French cities
12 month median 39.8
8hr max
Mean 23. 14
Median 46.9 ppb;
8-h average
IQR
35.9 to 43.7
Range
18.65 to 31. 17
25th-75th
41-52
See left
Mengetal. (2010):
Asthma ED visits or hospitalizations;
San Joaquin Valley, CA
Median 30.3 ppb
Yearly based on hourly
25-75% range
27.1 to 34.0
Moore et al. (2008):
Asthma hospital admissions;
South Coast Basin
Median 87.8 ppb
Quarterly 1 hr daily max
Range
28.6 to 199.9
Rage et al. (2009a):
Asthma severity;
Five French cities
Mean 30 ppb
8-h average
25th-75th
21-36
Wenten et al. (2009):
Respiratory school absence,
U.S.
Median 46.9 ppb;
10a.m. - 6 p.m. average
Min-Max
27.6-65.3
Long-term O3 exposure was related to first childhood asthma hospital admissions in
a positive concentration-response relationship in a New York State birth cohort (Lin
et al., 2008b). A separate hospitalization cross-sectional study in San Joaquin Valley,
California reports similar findings (Meng et al., 2010). Another study relates asthma
hospital admissions to quarterly average O3 in the South Coast Air Basin of
California (Moore et al., 2008).
7-33
-------
Information from toxicological studies indicates that long term exposure to O3 during
gestation or development can result in irreversible morphological changes in the
lung, which in turn can influence the function of the respiratory tract. Studies by
Plopper and colleagues using an allergic asthma model have demonstrated changes in
pulmonary function and airway morphology in adult and infant nonhuman primates
repeatedly exposed to environmentally relevant concentrations of O3 (Fanucchi et al..
2006: Joad et al.. 2006: Schelegle et al.. 2003: Harkemaet al.. 1987b). This
nonhuman primate evidence of an O3-induced change in airway responsiveness
supports the biologic plausibility of long term exposure to O3 contributing to effects
of asthma in children. Results from epidemiologic studies examining long-term O3
exposure and pulmonary function effects are inconclusive with some new studies
relating effects at higher exposure levels. The definitive 8-year follow-up analysis of
the first cohort of the CHS, which is discussed in Section 7.2 (Gauderman et al..
2004). provided little evidence that long-term exposure to ambient O3 was associated
with significant deficits in the growth rate of lung function in children. Other cross-
sectional studies provide mixed results.
Several studies (see Table 7-3) provide results adjusted for potential confounders,
presenting results for both O3 and PM (single and multipollutant models) as well as
other pollutants where PM effects were not provided. As shown in the table, O3
associations are generally robust to adjustment for potential confounding by PM.
7-34
-------
Table 7-3 Studies providing evidence concerning potential confounding by
PM for available endpoints.
Study and
Endpoint Exposure
Single
Pollutant O3
Single
Pollutant PM
O3 with PM
PM with O3
Asthma Related Health Effect Endpoint
Akinbamietal. (2010) IQR
Asthma prevalence in 35.9-43.7 ppb
children
Hwang et al. (2005) 10ppbO3
Asthma risk in children
Jacguemin et al. (2012) IQR 25-38 ppb
Asthma control in adults °* summer
Lee et al. (2009b) High O3 >50 ppb
Bronchitic symptoms
asthmatics
Lin et al. (2008b) IQR 2.5%
Asthma admissions in
children
Mena et al. (2007) 1 ppm
Asthma control
Mengetal. (2010) 1 0 ppb
Asthma ED visits,
Hospitalization
Rage et al. (2009b) IQR
Asthma severity in adults 28.5-33.9 ppb
1.56
(1.15,2.10)
1.138
(1 .001 , 1 .293
1.69
(1 .22, 2.34)
1.42
(0.75, 2.70)
1.16
(1.15, 1.17)
1.70
(0.91,3.18)
1.49
(1.05, 2.11)
2.53
(1.69, 3.79)
PM2.5
1.43
(0.98,2.10)
0.934
(0.909, 0.960)
1.33
(1.06, 1.67)
NA
NA
PM10
2.06
(1.17,3.61)
women
PM10
1.29
(0.99, 1 .69)
NA
Adjusted for
S02,PM25,PM10
1 .86 (1 .02-3.40)
Adjusted for
PM25, PM10
1.36(0.91-2.02)
PM10
1.253
(1 .089, 1 .442)
PM10
1.50
(1.07,2.11)
No substantial
differences
PM10, PM2.5
Air Quality Index
1.24
(1 .23, 1 .25)
Did not differ
Did not differ
No PM data
Three pollutant
(03, N02,S02)
2.74
(1.68,4.48)
PM2.5
1.24
(0.70-2.21)
PM2.5
1.26
0.80-1.98)
0.925
(0.899, 0.952)
1.28
(1 .06, 1 .55)
NA
NA
NA
NA
NA
Other Respiratory Health Effect Endpoints
Karretal. (2007) 1 0 ppb
Bronchiolitis
Hospitalization
Parker et al. (2009) 10 ppb
Respiratory allergy
Roias-Martinez et al. 11. 3 ppb IQR
(2007)
FEV, (ml) Deficit
Girls
0.92
(0.88, 0.96)
1.24
(1.15, 1.34)
-24
(-30, -1 9)
1.09
(1.04, 1.14)
1.23
(1 .04, 1 .46)
PM10IQR
36.4 ug/m3
-29(-36, -21)
PM2.5
1.02
(0.94, 1.10)
Multipollutant
1.18
(1 .09, 1 .27)
-17
(-23, -12)
1.09
(1.03, 1.15)
1.29(
1 .07, 1 .56)
-24
(-31 ,-16)
The highest quartile is shown for all results
NA = not available
7-35
-------
There is limited evidence for an association between long-term exposure to ambient
O3 concentrations and respiratory mortality (Jerrett et al., 2009) and this effect was
robust to the inclusion of PM2.5. The association between increased O3
concentrations and increased risk of death from respiratory causes was insensitive to
a number of different model specifications. Additionally, there is evidence that long-
term exposure to O3 is associated with mortality among individuals that had
previously experienced an emergency hospital admission due to COPD (Zanobetti
and Schwartz. 2011).
Taken together, the recent epidemiologic studies of respiratory health effects
(including respiratory symptoms, new-onset asthma and respiratory mortality)
combined with toxicological studies in rodents and nonhuman primates, provide
biologically plausible evidence that there is likely to be a causal relationship
between long-term exposure to O3 and respiratory effects. The epidemiologic
evidence includes studies that evaluate the relationship between long-term O3
exposure and respiratory effects such as studies that demonstrate interactions
between exercise or different genetic variants and long-term measures of O3
exposure on new-onset asthma in children; and increased respiratory symptom
effects in asthmatics. Additional studies of respiratory health effects and a study of
respiratory mortality provide a collective body of evidence supporting these
relationships. Studies considering other pollutants provide data suggesting that the
effects related to O3 are independent from potential effects of the other pollutants.
Some studies provide evidence for a positive concentration-response relationship.
Short-term studies provide supportive evidence with increases in respiratory
symptoms and asthma medication use, hospital admissions and ED visits for all
respiratory outcomes and asthma, and decrements in lung function in children.
The recent epidemiologic and toxicological data base provides a compelling case to
support the hypothesis that a relationship exists between long-term exposure to
ambient O3 and measures of respiratory health effects.
7.3 Cardiovascular Effects
7.3.1 Cardiovascular Disease
7.3.1.1 Cardiovascular Epidemiology
Long-term exposure to O3 and its effects on cardiovascular morbidity were not
considered in the 2006 O3 AQCD (U.S. EPA, 2006b). However, recent studies have
assessed the chronic effects of O3 concentration on cardiovascular morbidity
(Chuang et al., 2011; Forbes et al., 2009a; Chen et al., 2007a). The association
between O3 concentration and markers of lipid peroxidation and antioxidant capacity
was examined among 120 nonsmoking healthy college students, aged 18-22 years,
from the University of California, Berkeley (February—June 2002) (Chen et al.,
7-36
-------
2007a). By design, students were chosen from geographic areas so they had
experienced different concentrations of O3 over their lifetimes and during recent
summer vacation in either greater Los Angeles (LA) or the San Francisco Bay Area
(SF). A marker of lipid peroxidation, 8-isoprostane (8-iso-PGF) in plasma, was
assessed. This marker is formed continuously under normal physiological conditions
but has been found at elevated concentrations in response to environmental
exposures. A marker of overall antioxidant capacity, ferric reducing ability of plasma
(FRAP), was also measured. The lifetime average O3 concentration estimates (from
estimated monthly averages) did not show much overlap between the two geographic
areas [median (range): LA, 42.9 ppb (28.5-65.3); SF, 26.9 ppb (17.6-33.5)].
Estimated lifetime average O3 concentration was related to 8-iso-PGF [(3 = 0.025
(pg/mL)/8-h ppb O3, p = 0.0007]. For the 17-ppb lifetime O3 concentration
difference between LA and SF participants, there was a 17.41-pg/mL (95% CI:
15.43, 19.39) increase in 8-iso-PGF. No evidence of association was observed
between lifetime O3 concentration and FRAP [(3 = -2.21 (pg/mL)/8-h ppb O3,
p = 0.45]. The authors note that O3 was highly correlated with PMio_2.s and NO2 in
this study population; however, their inclusion in the O3 models did not substantially
modify the magnitude of the associations with O3. Because the average lifetime
concentration results were supported by shorter-term exposure period results from
analyses considering O3 concentrations up to 30 days prior to sampling, the authors
conclude that persistent exposure to O3 can lead to sustained oxidative stress and
increased lipid peroxidation. However, because there was not much overlap in
average lifetime O3 concentration estimates between LA and SF, it is possible that
the risk estimates involving the lifetime O3 exposures could be confounded by
unmeasured factors related to other differences between the two cities.
Forbes et al. (2009a) used the annual average exposures to assess the relationship
between chronic ambient air pollution and levels of fibrinogen and C-reactive protein
(CRP) in a cross-sectional study conducted in England. Data were collected from the
Health Survey of England for 1994, 1998, and 2003. The sampling strategy was
designed to obtain a representative sample of the English population; however, due
to small group sizes, only data from white ethnic groups were analyzed. For analyses,
the annual concentrations of O3 were averaged for the year of data collection and the
previous year with the exception of 1994 (because pollutant data were not available
for 1993). Median O3 concentrations were 26.7 ppb, 25.4 ppb, and 28 ppb for 1994,
1998, and 2003, respectively. Year specific adjusted effect estimates were created
and combined in a meta-analysis. No evidence of association was observed for O3
and levels of fibrinogen or CRP (e.g., the combined estimates for the percent change
in fibrinogen and CRP for a 10 ppb increase in O3 were -0.28 [95% CI: -2.43, 1.92]
and -3.05 [95% CI: -16.10, 12.02], respectively).
A study was performed in Taiwan to examine the association between long-term O3
concentrations and blood pressure and blood markers using the Social Environment
and Biomarkers of Aging Study (SEBAS) (Chuang et al., 2011). Individuals included
in the study were 54 years of age and older. The mean annual O3 concentration
during the study period was 22.95 ppb (SD 6.76 ppb). Positive associations were
observed between O3 concentrations and both systolic and diastolic blood pressure
7-37
-------
[changes in systolic and diastolic blood pressure were 21.51mmHg (95% CI: 16.90,
26.13) and 20.56 mmHg (95% CI: 18.14, 22.97) per 8.95 ppb increase in O3,
respectively]. Increased O3 concentrations were also associated with increased levels
of total cholesterol, fasting glucose, hemoglobin Ale, and neutrophils.
No associations were observed between O3 concentrations and triglyceride and IL-6
levels. The observed associations were reduced when other pollutants were added to
the models. Further research will be important for understanding the effects, if any,
of chronic O3 exposure on cardiovascular morbidity risk.
7.3.1.2 Cardiovascular Toxicology
Three new studies have investigated the cardiovascular effects of long-term exposure
to O3 in animal models (see Table 7-4 for study details). In addition to the short-term
exposure effects described in Section 6.3.3, a recent study found that O3 exposure in
genetically hyperlipidemic mice enhanced aortic atherosclerotic lesion area
compared to air exposed controls (Chuang et al., 2009). Chuang et al. (2009) not only
provided evidence for increased atherogenesis in susceptible mice, but also reported
an elevated vascular inflammatory and redox state in wild-type mice and infant
primates (Section 6.3.3). This study is compelling in that it identifies biochemical
and cellular events responsible for transducing the airway epithelial reactions of O3
into proinflammatory responses that are apparent in the extrapulmonary vasculature
(Cole and Freeman, 2009).
Another recent study provides further evidence for increased vascular inflammation
and oxidation and long term effects in the extrapulmonary space. Rats episodically
exposed to O3 for 16 weeks presented marked increases in gene expression of
biomarkers of oxidative stress, thrombosis, vasoconstriction, and proteolysis
(Kodavanti et al., 2011). Ozone exposure upregulated aortic mRNA expression of
heme oxygenase-1 (HO-1), tissue plasminogen activator (tPA), plasminogen
activator inhibitor-1 (PAI-1), von Willebrand factor (vWf), thrombomodulin,
endothelial nitric oxide synthase (eNOS), endothelin-1 (ET-1), matrix
metalloprotease-2 (MMP-2), matrix metalloprotease-3 (MMP-3), and tissue inhibitor
of matrix metalloprotease-2 (TIMP-2). In addition, O3 exposure depleted some
cardiac mitochondrial phospholipid fatty acids (C16:0 and C18:l), which may be the
result of oxidative modifications. The authors speculate that oxi datively modified
lipids and proteins produced in the lung and heart promote vascular pathology
through activation of lectin-like oxidized-low density lipoprotein receptor-1
(LOX-1). Activated LOX-1 induces expression of a number of the biomarkers
induced by O3 exposure and is considered pro-atherogenic. Both LOX-1 mRNA and
protein were increased in mouse aorta after O3 exposure. This study provides a
possible pathway and further support to the observed O3 induced atherosclerosis.
Vascular occlusion resulting from atherosclerosis can block blood flow through
vessels causing ischemia. The restoration of blood flow or reperfusion can cause
injury to the tissue from subsequent inflammation and oxidative damage. Ozone
7-38
-------
exposure enhanced the sensitivity to myocardial ischemia-reperfusion (I/R) injury in
rats while increasing oxidative stress levels and pro-inflammatory mediators and
decreasing production of anti-inflammatory proteins (Perepu et al., 2010). Both long-
and short-term O3 exposure decreased the left ventricular developed pressure, rate of
change of pressure development, and rate of change of pressure decay and increased
left ventricular end diastolic pressure in isolated perfused hearts (Section 6.3.3 for
short-term exposure discussion). In this ex vivo heart model, O3 induced oxidative
stress by decreasing SOD enzyme activity and increasing malondialdehyde levels.
Ozone also elicited a proinflammatory state evident by an increase in TNF-a and a
decrease in the anti-inflammatory cytokine IL-10. The authors conclude that O3
exposure will result in a greater I/R injury.
Overall, the few animal studies that have been conducted suggest that long-term O3
exposure may result in cardiovascular effects. These studies demonstrate O3-induced
atherosclerosis and injury. In addition, evidence is presented for a potential
mechanism for the development of vascular pathology that involves increased
oxidative stress and proinflammatory mediators, activation of LOX-1 by O3 oxidized
lipids and proteins, and upregulation of genes responsible for proteolysis,
thrombosis, and vasoconstriction. Further discussion of the mechanisms that may
lead to cardiovascular effects from O3 exposure can be found in Section 5.3.8.
Table 7-4
Study
Chuanq et al. (2009)
Kodavanti et al.
(2011)
Perepu etal. (2010)
Characterization of study details for Section 7.3.1.2.
Model
Mice; ApoE"'"; M;
6 weeks
Rat; Wistar; M;
10-12 weeks
Rat; Sprague-Dawley;
Weight: 50-75 g
Os (ppm) Exposure Duration
0.5 8 wks, 5 days/week,
8 h/day
0.4 16 wks, 1 day/week,
5 h/day
0.8 56 days, 8 h/day
Effects
Enhanced aortic atherosclerotic lesion
area compared to air controls.
Increased vascular inflammation and
oxidative stress, possibly through
activation of LOX-1 signaling.
Enhanced the sensitivity to myocardial I/R
injury while increasing oxidative stress
and pro-inflammatory mediators and
decreasing production of
anti-inflammatory proteins.
No previous studies investigated cardiovascular effects from long-term exposure to O3.
For details, see Section 7.3.1.2.
7.3.2 Cardiovascular Mortality
A limited number of epidemiologic studies have assessed the relationship between
long-term exposure to O3 and mortality. The 2006 O3 AQCD concluded that an
insufficient amount of evidence existed "to suggest a causal relationship between
chronic O3 exposure and increased risk for mortality in humans" (U.S. EPA, 2006b).
Though total and cardio-pulmonary mortality were considered in these studies,
cardiovascular mortality was not specifically considered. In the most recent follow-
up analysis of the ACS cohort (Jerrett et al., 2009), cardiopulmonary deaths were
7-39
-------
subdivided into respiratory and cardiovascular, separately, as opposed to combined in
the Pope et al. (2002) work. A 10-ppb increment in exposure to O3 elevated the risk
of death from the cardiopulmonary, cardiovascular, and ischemic heart disease.
Inclusion of PM2.5 as a copollutant attenuated the association with exposure to O3 for
all of the cardiovascular endpoints to become null. Additionally, a recent study
(Zanobetti and Schwartz. 2011) observed an association between long-term exposure
to O3 and elevated risk of mortality among Medicare enrollees that had previously
experienced an emergency hospital admission due to congestive heart failure (CHF)
or myocardial infarction (MI).
7.3.3 Summary and Causal Determination
Previous AQCDs did not address the cardiovascular effects of long-term O3 exposure
due to limited data availability. The evidence remains limited; however the emerging
data are supportive of a role for O3 in chronic cardiovascular diseases. Few
epidemiologic studies have investigated cardiovascular morbidity after long-term O3
exposure, and the majority only assessed cardiovascular disease related biomarkers.
The studies used annual or multi-year averages of air monitoring data for exposure
assessment. As described in Section 4.6, this exposure assignment method is typical
of long-term epidemiologic studies, and analyses suggest that annual average
concentrations are representative of exposure metrics accounting for residential
mobility. A study on O3 and cardiovascular mortality reported no association after
adjustment for PM2.s levels. Further epidemiologic studies on cardiovascular
morbidity and mortality after long-term exposure have not been published.
Toxicological evidence on long-term O3 exposure is also limited but three strong
toxicological studies have been published since the previous AQCD. These studies
provide evidence for O3 enhanced atherosclerosis and I/R injury, corresponding with
development of a systemic oxidative, proinflammatory environment. Further
discussion of the mechanisms that may lead to cardiovascular effects can be found in
Section 5.3.8. Although questions exist for how O3 inhalation causes systemic
effects, a recent study proposes a mechanism for development of vascular pathology
that involves activation of LOX-1 by O3 oxidized lipids and proteins. This activation
may also be responsible for O3 induced changes in genes involved in proteolysis,
thrombosis, and vasoconstriction. Taking into consideration the findings of
toxicological studies, and the emerging evidence from epidemiologic studies, the
generally limited body of evidence is suggestive of a causal relationship between
long-term exposures to O3 and cardiovascular effects.
7.4 Reproductive and Developmental Effects
Although the body of literature characterizing the health effects associated with
exposure to O3 is large and continues to grow, the research focusing on adverse birth
outcomes is relatively small. Among these studies, various measures of birth weight
7-40
-------
and fetal growth, such as low birth weight (LEW), small for gestational age (SGA),
and intrauterine growth restriction (IUGR), and preterm birth (<37-week gestation;
[PTB]) have received more attention in air pollution research, while congenital
malformations are less studied. There are also recent studies on reproductive and
developmental effects and infant mortality.
A major issue in studying environmental exposures and reproductive and
developmental effects (including infant mortality) is selecting the relevant exposure
period, since the biological mechanisms leading to these outcomes and the critical
periods of exposure are poorly understood. To account for this, many epidemiologic
studies evaluate multiple exposure periods, including long-term (months to years)
exposure periods, such as entire pregnancy, individual trimesters or months of
pregnancy, and short-term (days to weeks) exposure periods such as the days and
weeks immediately preceding birth. Due to the length of gestation in rodents
(18-24 days, on average), animal toxicological studies investigating the effects of O3
generally utilize short-term exposure periods. Thus, an epidemiologic study that uses
the entire pregnancy as the exposure period is considered to have a long-term
exposure period (about 40 weeks, on average), while a toxicological study conducted
with rats that also uses the entire pregnancy as the exposure period is considered to
have a short-term exposure period (about 18-24 days, on average). In order to
characterize the weight of evidence for the effects of O3 on reproductive and
developmental effects in a consistent, cohesive and integrated manner, results from
both short-term and long-term exposure periods are included in this section and are
identified accordingly in the text and tables throughout this section.
Due to the poorly understood biological mechanisms and uncertainty regarding
relevant exposure studies, all of the studies of reproductive and developmental
outcomes, including infant mortality, are evaluated in this section. Infant
development processes, much like fetal development processes, may be particularly
sensitive to O3-induced health effects. Exposures proximate to the effect may be
most relevant if exposure causes an acute effect. However, exposure occurring in
early life might affect critical growth and development, with results observable later
in the first year of life, or cumulative exposure during the first year of life may be the
most important determinant. In dealing with the uncertainties surrounding these
issues, studies have considered several exposure metrics based on different periods of
exposure, including both short- and long-term exposure periods. In the toxicological
literature,a challenge in interpreting data from studies that use very young murine
pups, is that pups can have differential exposure to O3 doses, versus their respective
dams, because of the physiology and behavior associated with the early postnatal
period. Namely, young pups tend to nuzzle close to their mothers and are often
housed in cages with litter used in nest formation. Both the dam's fur and the bedding
can absorb and react with O3, decreasing the dose that a young animal might receive.
The reproductive and developmental studies are characterized in this chapter, as they
contribute to the weight of evidence for an effect of O3 on reproductive and
developmental effects.
7-41
-------
Infants and fetal development processes may be particularly at-risk for O3-induced
health effects, and although the physical mechanisms are not fully understood,
several hypotheses have been proposed; these include: oxidative stress, systemic
inflammation, vascular dysfunction and impaired immune function (Section 5.3).
Study of these outcomes can be difficult given the need for detailed exposure data
and potential residential movement of mothers during pregnancy. Air pollution
epidemiologic studies reviewed in the 2006 O3 AQCD (U.S. EPA. 2006b) examined
impacts on birth-related endpoints, including intrauterine, perinatal, postneonatal,
and infant deaths; premature births; intrauterine growth retardation; very low birth
weight (weight <1,500 grams) and low birth weight (weight <2,500 grams); and birth
defects. However, in the limited number of studies that investigated O3, no
associations were found between O3 and birth outcomes, with the possible exception
of birth defects.
Several recent articles have reviewed methodological issues relating to the study of
outdoor air pollution and adverse birth outcomes (Chen et al., 2010a; Woodruff et al.,
2009: Ritz and Wilhelm. 2008: Slama et al.. 2008). Some of the key challenges to
interpretation of these study results include the difficulty in assessing exposure as
most studies use existing monitoring networks to estimate individual exposure to
ambient air pollution; the inability to control for potential confounders such as other
risk factors that affect birth outcomes (e.g., smoking); evaluating the exposure
window (e.g., trimester) of importance; and limited evidence on the physiological
mechanism of these effects (Ritz and Wilhelm, 2008; Slama et al., 2008).
Overall, the evidence for an association between exposure to ambient O3 and
reproductive and developmental outcomes is growing, yet remains relatively small.
Recently, an international collaboration was formed to better understand the
relationships between air pollution and adverse birth outcomes and to examine some
of these methodological issues through standardized parallel analyses in datasets
from different countries (Woodruff et al., 2010). Initial results from this collaboration
have examined PM and birth weight (Parker et al., 2011); work on O3 has not yet
been performed. Although early animal studies (Kavlock et al., 1980) found that
exposure to O3 in the late gestation of pregnancy in rats led to some abnormal
neurological and behavioral performances for neonates, to date human studies have
reported inconsistent results for the association of ambient O3 concentrations and
birth outcomes.
7.4.1 Effects on Sperm
A limited amount of research has been conducted to examine the association between
air pollution and male reproductive outcomes, specifically semen quality. To date,
the epidemiologic studies have considered various exposure durations before semen
collection that encompass either the entire period of spermatogenesis (i.e., 90 days)
or key periods of sperm development that correspond to epididymal storage,
development of sperm motility, and spermatogenesis. In an analysis conducted as
7-42
-------
part of the Teplice Program, 18-year-old men residing in the heavily polluted district
of Teplice in the Czech Republic were found to be at greater risk of having
abnormalities in sperm morphology and chromatin integrity than men of similar age
residing in Prachatice, a less polluted district (Selevan et al., 2000; Sram et al., 1999).
A follow-up longitudinal study conducted on a subset of the same men from Teplice
revealed associations between total episodic air pollution and abnormalities in sperm
chromatin (Rubes et al.. 2005). A limitation of these studies is that they did not
identify specific pollutants or their concentrations.
More recent epidemiologic studies conducted in the U.S. have also reported
associations between ambient air pollution and sperm quality for individual air
pollutants, including O3 and PM2.5. In a repeated measures study in Los Angeles,
CA, Sokol et al. (2006) reported a reduction in average sperm concentration during
three exposure windows (short-term exposures of 0-9, 10-14, and 70-90 days before
semen collection, as well as long-term exposures of 0-90 days before semen
collection) associated with high ambient levels of O3 in healthy sperm donors. This
effect persisted under a joint additive model for O3, CO, NO2 and PMi0. The authors
did not detect a reduction in sperm count. Hansen et al. (2010) investigated the effect
of exposure to O3 and PM2.s (using the same exposure windows used by Sokol et al.
(2006) on sperm quality in three southeastern counties (Wake County, NC; Shelby
County, TN; Galveston County, TX). Outcomes included sperm concentration and
count, morphology, DNA integrity and chromatin maturity. Overall, the authors
found both protective and adverse effects, although some results suggested adverse
effects on sperm concentration, count and morphology.
The biological mechanisms linking ambient air pollution to decreased sperm quality
have yet to be determined, though O3-induced oxidative stress, inflammatory
reactions, and the induction of the formation of circulating toxic species have been
suggested as possible mechanisms (see Section 5.3.8). Decremental effects on
testicular morphology have been demonstrated in a toxicological study with
histological evidence of O3-induced depletion of germ cells in testicular tissue and
decreased seminiferous tubule epithelial layer. Jedlinska-Krakowska et al. (2006)
demonstrated histopathological evidence of impaired spermatogenesis (round
spermatids/ spermatocytes, giant spermatid cells, and focal epithelial desquamation
with denudation to the basement membrane). The exposure protocol used five-
month-old adult rats exposed to O3 as adults (long-term exposure, 0.5 ppm, 5 h/day
for 50 days). This degeneration could be rescued by vitamin E administration,
indicating an antioxidant effect. Vitamin C administration had no effect at low doses
of ascorbic acid and exacerbated the O3-dependent damage at high doses, as would
be expected as vitamin C can be a radical generator instead of an antioxidant at
higher doses. In summary, this study provided toxicological evidence of impaired
spermatogenesis with O3 exposure that was rescued with certain antioxidant
supplementation.
7-43
-------
Overall, there is limited epidemiologic evidence for an association with O3
concentration and decreased sperm concentration. A recent toxicological study
provides limited evidence for a possible biological mechanism (histopathology
showing impaired spermatogenesis) for such an association.
7.4.2 Effects on Reproduction
Evidence suggests that exposure to air pollutants during pregnancy may be
associated with adverse birth outcomes, which has been attributed to the increased
sensitivity of the fetus due to physiologic immaturity. Gametes (i.e., ova and sperm)
may be even more at-risk, especially outside of the human body, as occurs with
assisted reproduction. Smokers require twice the number of in vitro fertilization
(IVF) attempts to conceive as non-smokers (Feichtinger et al., 1997), suggesting that
a preconception exposure can be harmful to pregnancy. A recent study used an
established national-scale, log-normal kriging method to spatially estimate daily
mean concentrations of criteria pollutants at addresses of women undergoing their
first IVF cycle and at their IVF labs from 2000 to 2007 in the northeastern U.S.
(Legro et al., 2010). Increasing O3 concentration at the patient's address during
ovulation induction (short-term exposure, ~12 days) was significantly associated
with an increased chance of live birth (OR =1.13, [95% CI: 1.05, 1.22] per 10 ppb
increase), but with decreased odds of live birth when exposed from embryo transfer
to live birth (long-term exposure, -200 days) (OR = 0.79, [95% CI: 0.69, 0.90] per
10 ppb increase). After controlling for NO2 in a copollutant model, however, O3 was
no longer significantly associated with IVF failure. The results of this study suggest
that short-term exposure to O3 during ovulation was beneficial (perhaps due to early
conditioning to O3), whereas long-term exposure to O3 (e.g., during gestation) was
detrimental, and reduced the likelihood of a live birth.
In most toxicological studies, reproductive success appears to be unaffected by O3
exposure. Nonetheless, one study has reported that 25% of the BALB/c mouse dams
in the highest O3 exposure group (1.2 ppm, short-term exposure GD9-18) did not
complete a successful pregnancy, a significant reduction (Sharkhuu et al., 2011).
Ozone administration (continuous 0.4, 0.8 or 1.2 ppm O3) to CD-I mouse dams
during the majority of pregnancy (short-term exposure, PD7-17, which excludes the
pre-implantation period), led to no adverse effects on reproductive success
(proportion of successful pregnancies, litter size, sex ratio, frequency of still birth, or
neonatal mortality) (Bignami et al., 1994). There was a nearly statistically significant
increase in pregnancy duration (0.8 and 1.2 ppm O3). Initially, dam body weight (0.8
and 1.2 ppm O3), water consumption (0.4, 0.8 and 1.2 ppm O3) and food
consumption (0.4, 0.8 and 1.2 ppm O3) during pregnancy were decreased with O3
exposure but these deficits dissipated a week or two after the initial exposure
(Bignami et al., 1994). The anorexigenic effect of O3 exposure on the pregnant dam
appears to dissipate with time; the dams seem to adapt to the O3 exposure. In males,
data exist showing morphological evidence of altered spermatogenesis in O3 exposed
animals (Jedlinska-Krakowska et al., 2006). Some evidence suggests that O3 may
7-44
-------
affect reproductive success when combined with other chemicals. Kavlock et al.
(1979) showed that O3 acted synergistically with sodium salicylate to increase the
rate of pup resorptions after midgestational exposure (1.0 ppm O3, short-term
exposure, GD9-GD12). At low concentrations of O3 exposure, toxicological studies
show reproductive effects to include a transient anorexigenic effect of O3 on
gestational weight gain, and a synergistic effect of O3 on salicylate-induced pup
resorptions; other fecundity, pregnancy- and gestation-related outcomes appear
unaffected by O3 exposure.
Collectively, there is very little epidemiologic evidence for the effect of short- or
long-term exposure to O3 on reproductive success, and the reproductive success in
rats appears to be unaffected in toxicological studies of short-term exposure to O3.
7.4.3 Birth Weight
With birth weight routinely collected in vital statistics and being a powerful predictor
of infant mortality, it is the most studied outcome within air pollution-birth outcome
research. Air pollution researchers have analyzed birth weight as a continuous
variable and/or as a dichotomized variable in the form of LEW (<2,500 g [5 Ibs,
8 oz]).
Birth weight is primarily determined by gestational age and intrauterine growth, but
also depends on maternal, placental and fetal factors as well as on environmental
influences. In both developed and developing countries, LEW is the most important
predictor for neonatal mortality and is a significant determinant of postneonatal
mortality and morbidity. Studies report that infants who are smallest at birth have a
higher incidence of diseases and disabilities, which continue into adulthood (Hack
andFanaroff. 1999).
The strongest evidence for an effect of O3 on birth weight comes from the Children's
Health Study conducted in southern California. In this study, Salam et al. (2005)
report that maternal exposure to 24-h avg O3 concentrations averaged over the entire
pregnancy was associated with reduced birth weight (39.3 g decrease [95% CI: -55.8,
-22.8] in birth weight per 10 ppb and 8-h avg (19.2-g decrease [95% CI: -27.7, -10.7]
in birth weight per 10 ppb). This effect was stronger for concentrations averaged over
the second and third trimesters. PMi0, NO2 and CO concentrations averaged over the
entire pregnancy were not statistically significantly associated with birth weight,
although CO concentrations in the first trimester and PMi0 concentrations in the third
trimester were associated with a decrease in birth weight. Additionally, the authors
observed a concentration-response relationship of birth weight with 24-h avg O3
concentrations averaged over the entire pregnancy that was clearest above the 30-ppb
level (see Figure 7-4). Relative to the lowest decile of 24-h avg O3, estimates for the
next 5 lowest deciles were approximately -40 g to -50 g, with no clear trend and with
95% confidence bounds that included zero. The highest four deciles of O3 exposure
showed an approximately linear decrease in birth weight, and all four 95% CIs
7-45
-------
excluded zero, and ranged from mean decreases of 74 grams to decreases of
148 grams.
50
0
—, -50
'5 -100
° -150
-200
-250
Q
0
9 v 6
0
o
o
20 30 40
24-hr 03(ppb)
50
Note: Deficits are plotted against the decile-group-specific median O3 exposure. Error bars represent 95% CIs. Indicator variables
for each decile of O3 exposure (except the least-exposed group) were included in a mixed model.
Source: Salam et al. (2005).'
Figure 7-4 Birthweight deficit by decile of 24-h avg Os concentration averaged
over the entire pregnancy compared with the decile group with the
lowest Oz exposure.
Several additional studies conducted in the U.S. and Canada also investigated the
association between ambient O3 concentrations and birth weight and report some
weak evidence for an association. Morello-Frosch et al. (2010) estimated ambient O3
concentrations throughout pregnancy and for each trimester in the neighborhoods of
women who delivered term singleton births between 1996 and 2006 in California.
A 10-ppb increase in the O3 concentration averaged across the entire pregnancy was
associated with a 5.7-g decrease (95% CI: -6.6, -4.9) in birth weight when exposures
were calculated using monitors within 10 km of the maternal address at date of birth.
When the distance from the monitor was restricted to 3 km, the decrease in birth
weight associated with a 10-ppb increase in O3 concentration was 8.9 g (95% CI:
-10.6, -7.1). These results persisted in copollutant models and in models that
stratified by trimester of exposure, SES, and race. Darrow et al. (201 Ib) did not
observe an association with birth weight and O3 concentrations during two exposure
periods of interest (i.e., the first month and last trimester), but did find an association
7-46
-------
with reduced birth weight when examining the cumulative air pollution concentration
during the entire pregnancy period. Additionally, they observed effect modification
by race and ethnicity, such that associations between birth weight and third-trimester
O3 concentrations were significantly stronger in Hispanics and non-Hispanic African
Americans than in non-Hispanic whites. Chen et al. (2002) used 8-h avg O3
concentrations to create exposure variables based on average maternal exposure for
each trimester. Ozone was not found to be related to birth weight in single-pollutant
models, though the O3 effect during the third trimester was borderline statistically
significant in a copollutant model with PMi0.
Several studies found no association between ambient O3 concentrations and birth
weight. Wilhelm and Ritz (2005) extended previous analyses of term LEW (Ritz et
al.. 2000: Ritz and Yu. 1999) to include the period 1994-2000. The authors examined
varying residential distances from monitoring stations to see if the distance affected
risk estimation, exploring the possibility that effect attenuation may result from local
pollutant heterogeneity inadequately captured by ambient monitors. As in their
previous studies, the authors observed associations between elevated concentrations
of CO and PMi0 both early and late in pregnancy and risk of term LEW. After
adjusting for CO and/or PMi0 the authors did not observe associations between O3
and term LEW in any of their models. Brauer et al. (2008) evaluated the impacts of
air pollution (CO, NO2, NO, O3, SO2, PM2.5, PMi0) on birth weight for the period
1999-2002 using spatiotemporal residential exposure metrics by month of pregnancy
in Vancouver, EC. Quantitative results were not presented for the association
between O3 and LEW, though the authors observed associations that were largely
protective. Dugandzic et al. (2006) examined the association between LEW and
ambient levels of air pollutants by trimester of exposure among a cohort of term
singleton births from 1988-2000. Though there was some indication of an association
with SO2 and PMi0, there were no effects for O3.
Similarly, studies conducted in Australia, Latin America, and Asia report limited
evidence for an association between ambient O3 and measures of birth weight.
In Sydney, Australia, Mannes et al. (2005) found that O3 concentrations in the
second trimester of pregnancy had small adverse effects on birth weight (7.5-g
decrease; [95% CI: -13.8, 1.2] per 10 ppb), although this effect disappeared when the
analysis was limited to births with a maternal address within 5 km of a monitoring
station (87.7-g increase; [95% CI: 10.5, 164.9] per 10 ppb). Hansen et al. (2007)
reported that trimester and monthly specific exposures to all pollutants were not
statistically significantly associated with a reduction in birth weight in Brisbane,
Australia. In Sao Paulo, Brazil, Gouveia et al. (2004) found that O3 exhibited a small
inverse relation with birth weight over the third trimester (6.0-g decrease; [95% CI:
-30.8, 18.8] per 10 ppb). Lin et al. (2004b) reported a positive, though not
statistically significant, exposure-response relationship for O3 during the entire
pregnancy in a Taiwanese study. In a study performed in Korea, Ha et al. (2001)
reported no O3 effect during the first trimester of pregnancy, but they found that
during the third trimester of pregnancy O3 was associated with LEW (RR = 1.05
[95% CI: 1.02, 1.08] per 10 ppb).
7-47
-------
Table 7-5
Study
Salam et al. (2005)
Morello-Frosch et al.
(2010)
Darrowetal. (201 1b)
Chen et al. (2002)
Wilhelm and Ritz
(2005)
Brauer et al. (2008)
Dugandzic et al.
(2006)
Manneset al. (2005)
Hansen et al. (2007)
Gouveia et al. (2004)
Brief summary
Location
Sample Size
California, U.S.
(n = 3,901)
California, U.S.
(n = 3,545, 177)
Atlanta, GA
(N=406,627)
Northern Nevada,
U.S.
(n = 36,305)
Los Angeles County,
CA
(n = 136,134)
Vancouver, BC,
Canada
(n = 70,249)
Nova Scotia, Canada
(n = 74,284)
Sydney, Australia
(n = 138,056)
Brisbane, Australia
(n = 26,617)
Sao Paulo, Brazil
(n = 179,460)
of epidemiologic studies of birth weight.
Mean Os (ppb)
24-h avg:
27.3
8h:
50.6
24-h avg:
23.5
8-h max:
44.8
8-h:
27.2
1-h:
21.1-22.2
24-h avg:
14
24-h avg:
21
1-h max:
31.6
8 h max:
26.7
1-h max:
31.5
Exposure assessment
ZIP code level
Nearest Monitor
(within 10, 5, 3 km)
Population-weighted
spatial average
County level
Varying distances from
monitor
Nearest Monitor
(within 10 km)
Inverse Distance
Weighting (IDW)
Nearest Monitor
(within 25 km)
Citywide avg and
<5 km from monitor
Citywide avg
Citywide avg
Effect Estimate3
(95% Cl)
Entire pregnancy:
-39.3 g (-55.8, -22.8)
T1: -6.1 g (-16.8, 4.8)
T2: -20.0 g (-31 .7, -8.4)
T3: -20.7 g (-32.1, -9.3)
Entire pregnancy:
-5.7 g (-6.6, -4.9)
T1:-2.1 g (-2.9, -1.4)
T2: -2.3 g (-3.1, -1.5)
T3:-1.3g (-2.1, -0.6)
Entire pregnancy:
-12.3g (-17.8, -6.8)
First 28 days
-0.5 g (-3.0, 2.1)
T3: -0.9g (-4.5, 2.8)
Entire pregnancy:
20.9 g (6.3, 35.5)
T1 : 23.4 g (-35.6, 82.4)
T2: -1 9.4 g (-77.0, 38.2)
T3: 7.7 g (-50.9, 66.3)
T1: NR
T3: NR
6 weeks before birth: NR
Entire pregnancy: NR
First 30 days of pregnancy: NR
Last 30 days of pregnancy: NR
T1:NR
T3: NR
T1: 0.97(0.81, 1.18)d
T2: 1 .06 (0.87, 1 .27)d
T3:1.01 (0.83-1 .24)d
T1 : -0.9 g (-6.6, 4.8)
T2: -7.5 g (-13.8, 1.2)
T3:-4.5g (-10.8, 1.8)
Last 30 days:
-1.1 g (-5.6, 3.4)
T1 : 2.8 g (-10.5, 16.0)
T2:4.4g (-11.4, 20.1)
T3: 11. 3 g (-4.4, 27.1)
T1:-3.2g (-25.6, 19)
T2: -0.2 g (-23.8, 23.4)
T3:-6.0g (-30.8, -18.8)
7-48
-------
Study
Lin et al. (2004b)
Haetal. (2001)
Location
Sample Size
Kaohsiung and
Taipei, Taiwan
(n = 92,288)
Seoul, Korea
(n = 276,763)
Mean Os (ppb)
24-havg: 15.86-
47.78
8-h avg:
22.4-23.3"
Exposure assessment
Nearest monitor
(within 3 km)
Citywide avg
Effect Estimate3
(95% Cl)
Entire pregnancy:
1.13(0.92, 1.38)°
T1: 1.02 (0.85, 1.22)°
T2: 0.93 (0.78, 1.12)°
T3: 1 .05 (0.87, 1 .26)°
T1: 0.87 (0.81, 0.94)°
T3: 1.05(1.02, 1.08)°
"Change in birthweight per 10 ppb change in O3
"Median
°Odds ratios of LEW; Highest quartile of exposure compared to lowest quartile of exposure
dRelative risk of LEW per 10 ppb change in O3
T1 = First Trimester, T2 = Second Trimester, T3 = Third Trimester
NR: No quantitative results reported
Table 7-5 provides a brief overview of the epidemiologic studies of birth weight.
In summary, only the Children's Health Study conducted in southern California
(Salam et al., 2005) provides strong evidence for an effect of ambient O3 on birth
weight. The study by Morello-Frosch et al. (2010), also conducted in California,
provides support for the results of the Children's Health Study. Additional studies,
conducted in the U.S., Canada, Australia, Latin America, and Asia, provide limited
and inconsistent evidence to support the effect reported in the Children's Health
Study. The toxicological literature on the effect of O3 on birth weight is sparse.
In some studies, the reporting of birth weight may be avoided because birth weight
can be confounded by decreased litter size resulting from an increased rate of pup
resorption (aborted pups) in O3 exposed dams. In one toxicological study by Haro
andPaz (1993), no differences in litter size were observed and decreased birth weight
in pups from dams who were exposed to Ippm O3 during pregnancy (short-term
exposure, -22 days) was reported. A second animal toxicology study recapitulated
these finding with pregnant BALB/c mice exposed to O3 (1.2 ppm, short-term
exposure, GD9-18) that produced pups with significantly decreased birth weights
(Sharkhuuetal..2011).
7.4.4 Preterm Birth
Preterm birth (PTB) is a syndrome (Romero et al., 2006) that is characterized by
multiple etiologies. It is therefore unusual to be able to identify an exact cause for
each PTB. In addition, PTB is not an adverse outcome in itself, but an important
determinant of health status (i.e., neonatal morbidity and mortality). Although some
overlap exists for common risk factors, different etiologic entities related to distinct
risk factor profiles and leading to different neonatal and postneonatal complications
are attributed to PTB and measures of fetal growth. Although both restricted fetal
growth and PTB can result in LBW, prematurity does not have to result in LBW or
growth restricted babies.
7-49
-------
A major issue in studying environmental exposures and PTB is selecting the relevant
exposure period, since the biological mechanisms leading to PTB and the critical
periods of vulnerability are poorly understood (Bobak, 2000). Short-term exposures
proximate to the birth may be most relevant if exposure causes an acute effect.
However, exposure occurring in early gestation might affect placentation, with
results observable later in pregnancy, or cumulative exposure during pregnancy may
be the most important determinant. The studies reviewed have dealt with this issue in
different ways. Many have considered several exposure metrics based on different
periods of exposure. Often the time periods used are the first month (or first
trimester) of pregnancy and the last month (or 6 weeks) prior to delivery. Using a
time interval prior to delivery introduces an additional problem since cases and
controls are not in the same stage of development when they are compared. For
example, a preterm infant delivered at 36 weeks is a 32-week fetus 4 weeks prior to
birth, while an infant born at term (40 weeks) is a 36-week fetus 4 weeks prior to
birth.
Recently, investigators have examined the association of PTB with both short-term
(i.e., hours, days, or weeks) and long-term (i.e., months or years) exposure periods.
Time-series studies have been used to examine the association between air pollution
concentrations during the days immediately preceding birth. An advantage of these
time-series studies is that this approach can remove the influence of covariates that
vary across individuals over a short period of time. Retrospective cohort and case-
control studies have been used to examine long-term exposure periods, often
averaging air pollution concentrations over months or trimesters of pregnancy.
Studies of PTB fail to show consistency in pollutants and periods during pregnancy
when an effect occurs. For example, while some studies find the strongest effects
associated with exposures early in pregnancy, others report effects when the
exposure is limited to the second or third trimester. However, the effect of air
pollutant exposure during pregnancy on PTB has a biological basis. There is an
expanding list of possible mechanisms that may explain the association between O3
exposure and PTB (see Section 5.4.2.4).
Many studies of PTB compare exposure in quartiles, using the lowest quartile as the
reference (or control) group. No studies use a truly unexposed control group.
If exposure in the lowest quartile confers risk, than it may be difficult to demonstrate
additional risk associated with a higher quartile. Thus negative studies must be
interpreted with caution.
Preterm birth occurs both naturally (idiopathic PTB), and as a result of medical
intervention (iatrogenic PTB). Ritz et al. (2007); (2000) excluded all births by
Cesarean section to limit their studies to idiopathic PTB. No other studies attempted
to distinguish the type of PTB, although air pollution exposure maybe associated
with only one type. This is a source of potential effect misclassification.
Generally, studies of air pollution and birth outcomes conducted in North America
and the United Kingdom have not identified an association between PTB and
maternal exposure to O3. Most recently, Darrow et al. (2009) used vital record data
7-50
-------
to construct a retrospective cohort of 476,489 births occurring between 1994 and
2004 in 5 central counties of metropolitan Atlanta, GA. Using a time-series
approach, the authors examined aggregated daily counts of PTB in relation to
ambient levels of CO, NO2, SO2, O3, PMi0, PM2.5 and speciated PM measurements.
This study investigated 3 gestational windows of short- and long-term exposure: the
final week of gestation (short-term exposure), and the first month of gestation and the
final 6 weeks of gestation (long-term exposure). The authors did not observe
associations of PTB with O3 concentrations for any of the exposure periods.
A number of U.S. studies were conducted in southern California, and report
somewhat inconsistent results. Ritz et al. (2000) evaluated the effect of air pollution
(CO, NO2, O3, PM10) exposure during pregnancy on the occurrence of PTB in a
cohort of 97,518 neonates born in southern California between 1989 and 1993.
The authors use both short- and long-term exposure windows, averaging pollutant
measures taken at the closest air-monitoring station over distinct periods, such as 1,
2, 4, 6, 8, 12, and 26 weeks before birth and the whole pregnancy period.
Additionally, they calculated average exposures for the first and second months of
pregnancy. The authors found no consistent effects associated with O3 concentration
over any of the pregnancy periods in single or multipollutant models. Wilhelm and
Ritz (2005) extended previous analyses of PTB (Ritz et al.. 2000: Ritz and Yu. 1999)
in California to include 1994-2000. The authors examined varying residential
distances from monitoring stations to see if the distance affected risk estimation,
because effect attenuation may result from local pollutant heterogeneity inadequately
captured by ambient monitors. The authors analyzed the association between long-
term O3 exposure during varying periods of pregnancy and PTB, finding a positive
association between O3 levels in both the first trimester of pregnancy (RR = 1.23
[95% CI: 1.06, 1.42] per 10 ppb increase in 24-h avg O3) and the first month of
pregnancy (results for first trimester exposure were similar, but slightly smaller,
quantitative results not presented) in models containing all pollutants. No association
was observed between O3 in the 6 weeks before birth and preterm delivery. Finally,
Ritz et al. (2007) conducted a case-control survey nested within a birth cohort and
assessed the extent to which residual confounding and exposure misclassification
impacted air pollution effect estimates. The authors calculated mean long-term
exposure levels for three gestational periods: the entire pregnancy, the first trimester,
and the last 6 weeks before delivery. Though positive associations were observed for
CO and PM2 5, no consistent patterns of increase in the odds of PTB for O3 or NO2
were observed.
A study conducted in Canada evaluated the impacts of air pollution (including CO,
NO2, NO, O3, SO2, PM2.5, and PM10) on PTBs (1999-2002) using spatiotemporal
residential exposure metrics by month of pregnancy (long-term exposure) in
Vancouver, BC (Brauer et al., 2008). The authors did not observe consistent
associations with any of the pregnancy average exposure metrics except for PM2 5 for
PTB. The O3 associations were largely protective, and no quantitative results were
presented for O3. Additionally, Lee et al. (2008c) used time-series techniques to
investigate the associations of short-term exposure to O3 and PTB in London,
England. In addition to exposure on the day of birth, cumulative exposure up to
7-51
-------
1 week before birth was investigated. The risk of PTB did not increase with exposure
to the levels of ambient air pollution experienced by this population.
Conversely, studies conducted in Australia and China provide evidence for an
association between ambient O3 and PTB. Hansen et al. (2006) reported that long-
term exposure to O3 during the first trimester was associated with an increased risk
of PTB (OR = 1.38, [95% CI: 1.14, 1.69] per 10 ppb increase). Although the test for
trend was significant due to the strong effect in the highest quartile, there was not an
obvious exposure-response pattern across the quartiles of O3 during the first
trimester. The effect estimate was diminished and lost statistical significance when
PMi0 was included in the model (OR = 1.23, [95% CI: 0.97, 1.59] per 10 ppb
increase). Maternal exposure to O3 during the 90 days prior to birth showed a weak,
positive association with PTB (OR = 1.09, [95% CI: 0.85, 1.39] per 10 ppb increase).
Jalaludin et al. (2007) found that O3 levels in the month and three months preceding
birth had a statistically significant association with PTB. Ozone levels in the first
trimester of pregnancy were associated with increased risks for PTBs (OR =1.15
[95% CI: 1.05, 1.24] per 10 ppb increase in 1-h max O3 concentration), and remained
a significant predictor of PTB in copollutant models (ORs between 1.07 and 1.10).
Jiang et al. (2007) examined the effect of short- and long-term exposure to air
pollution on PTB, including risk in relation to levels of pollutants for a single day
exposure window with lags from 0 to 6 days before birth. An increase of 10 ppb of
the 8-week avg of O3 corresponded to 9.47% (95% CI: 0.70, 18.7%) increase in
PTBs. Increases in PTB were also observed for PMi0, SO2, and NO2. The authors
did not observe any significant effect of short-term exposure to outdoor air pollution
on PTB among the 1 -day time windows examined in the week before birth.
Few data are available from toxicological studies; a study reported a nearly
statistically significant increase in pregnancy duration (short-term exposure) in mice
when exposed to 0.8 or 1.2 ppm O3. This phenomenon was most likely due to the
anorexigenic effect of relatively high O3 concentrations (Bignami et al., 1994).
Table 7-6 provides a brief overview of the epidemiologic studies of PTB.
In summary, the evidence is consistent when examining short-term exposure to O3
during late pregnancy and reports no association with PTB. However when long-term
exposure to O3 early in pregnancy is examined the results are inconsistent.
Generally, studies conducted in the U.S., Canada, and England find no association
with O3 and PTB, while studies conducted in Australia and China report an O3 effect
on PTB.
7-52
-------
Table 7-6 Brief summary of epidemiologic studies of preterm birth (PTB).
Study
Darrow et al. (2009)
Ritz et al. (2000)
Wilhelm and Ritz
(2005)
Ritz et al. (2007)
Brauer et al. (2008)
Lee et al. (2008c)
Hansen et al. (2006)
Jalaludin et al.
(2007)
Jiang et al. (2007)
Location
Sample Size
Atlanta, GA
(n = 476,489)
California, U.S.
(n = 97,158)
Los Angeles,
CA
(n = 106,483)
Los Angeles,
CA
(n = 58,31 6)
Vancouver, BC,
Canada
(n = 70,249)
London, UK
Brisbane,
Australia
(n = 28,200)
Sydney,
Australia
(n = 123,840)
Shanghai,
China
(n = 3,346
preterm births)
Mean O3
(PPb)
8-h max:
44.1
8h:
36.9
1 h:
21.1-22.2
24-h avg:
22.5
24-h avg:
14
24-h avg:
NR
8-h max:
26.7
1-h max:
30.9
8-h avg:
32.7
Exposure
assessment
Population-
weighted spatial
averages
Nearest Monitor
(within 4 miles)
<2 mi of monitor
Varying distances
to monitor
Nearest monitor to
ZIP code
Nearest Monitor
(within 10 km)
Inverse Distance
Weighting (IDW)
1 monitor
Citywide avg
Citywide avg
and <5 km from
monitor
Citywide avg
Effect Estimate3 (95% Cl)
First month: 0.98 (0.97, 1 .00)
Last week: 0.99 (0.98, 1 .00)
Last 6 weeks: 1 .00 (0.98, 1 .02)
First month: NR
Last 6 weeks: NR
First month: 1 .23 (1 .06, 1 .42)
T1: NR
T2: 1.38(1.14, 1.66)
Last 6 weeks: NR
Entire pregnancy: NR
T1 : 0.93 (0.82, 1 .06)
Last 6 weeks: NR
Entire pregnancy: NR
First 30 days of pregnancy: NR
Last 30 days of pregnancy: NR
T1: NR
T3: NR
LagO: 1.00(1.00, 1.01)
T1: 1.39(1.15, 1.70)
T3: 1 .09 (0.88, 1 .39)
First month: 1.04(0.95, 1.13)
T1: 1.15(1.05,1.24)
T3: 0.98 (0.89, 1 .07)
Last month: 0.98 (0.88, 1 .06)
4 wks before birth: 1 .06 (1 .00, 1.12)
6 wks before birth: 1 .06 (0.99, 1.13)
8 wks before birth: 1 .09 (1 .01,1.19)
Acute effects, LO to L6:
NR* (relative risk results presented in figure 1)
*NR: No quantitative results reported; however,
preterm birth was not significantly associated with
outdoor O3 air pollution in any lag day (0 - 6) that
was considered in the Jiana et al. (2007) study.
aRelative risk of PTB per 10 ppb change in O3.
T1 = First Trimester, T2 = Second Trimester, T3 = Third Trimester.
LO = Lag 0, L1 = Lag 1, L2 = Lag 2, L3 = Lag 3, L4 = Lag 4, L5 = Lag 5, L6 = Lag 6.
NR: No quantitative results reported.
7-53
-------
7.4.5 Fetal Growth
Low birth weight has often been used as an outcome measure because it is easily
available and accurately recorded on birth certificates. However, LEW may result
from either short gestation, or inadequate growth in utero. Most of the studies
investigating air pollution exposure and LEW limited their analyses to term infants to
focus on inadequate growth. A number of studies were identified that specifically
addressed growth restriction in utero by identifying infants who failed to meet
specific growth standards. Usually these infants had birth weight less than the 10th
percentile for gestational age, using an external standard. Many of these studies have
been previously discussed, since they also examined other reproductive outcomes
(i.e.,LBWorPTB).
Fetal growth is influenced by maternal, placental, and fetal factors. The biological
mechanisms by which air pollutants may influence the developing fetus remain
largely unknown. Several mechanisms have been proposed, and are the same as those
hypothesized for birth weight (see Section 5.4.2.4). Additionally, in animal
toxicology studies, O3 causes transient anorexia in exposed pregnant dams. This may
be one of many possible contributors to O3-dependent decreased fetal growth.
A limitation of environmental studies that use birth weight as a proxy measure of
fetal growth is that patterns of fetal growth during pregnancy cannot be assessed.
This is particularly important when investigating pollutant exposures during early
pregnancy as birth weight is recorded many months after the exposure period.
The insult of air pollution may have a transient effect on fetal growth, where growth
is hindered at one point in time but catches up at a later point. For example, maternal
smoking during pregnancy can alter the growth rate of individual body segments of
the fetus at variable developmental stages, as the fetus experiences selective growth
restriction and augmentation (Lampl and Jeantv. 2003).
The terms small-for-gestational-age (SGA), which is defined as a birth weight <10th
percentile for gestational age (and often sex and/or race), and intrauterine growth
retardation (IUGR) are often used interchangeably. However, this definition of SGA
does have limitations. For example, using it for IUGR may overestimate the
percentage of "growth-restricted" neonates as it is unlikely that 10% of neonates
have growth restriction (Wollmann, 1998). On the other hand, when the 10th
percentile is based on the distribution of live births at a population level, the
percentage of SGA among PTB is most likely underestimated (Hutcheon and Platt
2008). Nevertheless, SGA represents a statistical description of a small neonate,
whereas the term IUGR is reserved for those with clinical evidence of abnormal
growth. Thus all IUGR neonates will be SGA, but not all SGA neonates with be
IUGR (Wollmann. 1998). In the following section the terms SGA and IUGR are
referred to as each cited study used the terms.
Over the past decade a number of studies examined various metrics of fetal growth
restriction. Salam et al. (2005) assessed the effect of increasing O3 concentrations on
IUGR in a population of infants born in California from 1975-1987 as part of the
7-54
-------
Children's Health Study. The authors reported that maternal O3 exposures averaged
over the entire pregnancy and during the third trimester were associated with
increased risk of IUGR. A 10-ppb difference in 24-h maternal O3 exposure during
the third trimester increased the risk of IUGR by 11% (95% CI: 0, 20%). Brauer et
al. (2008) evaluated the impacts of air pollution (CO, NO2, NO, O3, SO2, PM2.5,
PMio) on SGA (1999-2002) using spatiotemporal residential exposure metrics by
month of pregnancy in Vancouver, BC. The O3 associations were largely protective
(OR = 0.87, [95% CI: 0.81, 0.93] for a 10 ppb increase in inverse distance weighted
SGA), and no additional quantitative results were presented for O3. Liu et al. (2007b)
examined the association between IUGR among singleton term live births and SO2,
NO2, CO, O3, and PM2 5 in 3 Canadian cities for the period 1985-2000. No increase
in the risk of IUGR in relation to exposure to O3 averaged over each month and
trimester of pregnancy was noted.
Three studies conducted in Australia provide evidence for an association between
ambient O3 and fetal growth restriction. Hansen et al. (2007) examined SGA among
singleton, full-term births in Brisbane, Australia in relation to ambient air pollution
(bsp, PMio, NO2, O3) during pregnancy. They also examined head circumference
and crown-heel length in a subsample of term neonates. Trimester specific exposures
to all pollutants were not statistically significantly associated with a reduction in head
circumference or an increased risk of SGA. When monthly-specific exposures were
examined, the authors observed an increased risk of SGA associated with exposure to
O3 during month 4 (OR =1.11 [95% CI: 1.00, 1.24] per 10 ppb increase). In a
subsequent study, Hansen et al. (2008) examined the possible associations between
fetal ultrasonic measurements and ambient air pollution (PMio, O3, NO2, SO2)
during early pregnancy. This study had two strengths: (1) fetal growth was assessed
during pregnancy as opposed to at birth; and (2) there was little delay between
exposures and fetal growth measurements, which reduces potential confounding and
uses exposures that are concurrent with the observed growth pattern of the fetus.
Fetal ultrasound biometric measurements were recorded for biparietal diameter
(BPD), femur length, abdominal circumference, and head circumference. To further
improve exposure assessment, the authors restricted the samples to include only
scans from women for whom the centroid of their postcode was within 14 km of an
air pollution monitoring site. Ozone during days 31 -60 was associated with decreases
in all of the fetal growth measurements, and a 1.78 mm reduction in abdomen
circumference per 10 ppb increase in O3 concentration, though this effect did not
persist in copollutant models. The change in ultrasound measurements associated
with O3 during days 31-60 of gestation indicated that increasing O3 concentration
decreased the magnitude of ultrasound measurements for women living within 2 km
of the monitoring site. The relationship decreased toward the null as the distance
from the monitoring sites increased. When assessing effect modification due to SES,
there was some evidence of effect modification for most of the associations, with the
effects of air pollution stronger in the highest SES quartile. In the third study,
Mannes et al. (2005) estimated the effects of pollutant (PMi0, PM2.5, NO2, CO and
O3) exposure in the first, second and third trimesters of pregnancy and risk of SGA
in Sydney, Australia. Citywide average air pollutant concentrations in the last month,
third trimester, and first trimester of pregnancy had no effect on SGA.
7-55
-------
Concentrations of O3 in the second trimester of pregnancy had small but adverse
effects on SGA (OR =1.10 [95% CI: 1.00, 1.14] per 10 ppb increment). This effect
disappeared when the analysis was limited to births with a maternal address within 5
km of a monitoring station (OR = 1.00 [95% CI: 0.60, 1.79] per 10 ppb increment).
Very little information from toxicological studies is available to address effects on
fetal growth. However, there is evidence to suggest that prenatal (short-term)
exposure to O3 can affect postnatal growth. A few studies reported that mice or rats
exposed developmentally (gestationally ± lactationally) to O3 had deficits in body
weight gain in the postpartum period (Bignami et al., 1994; Haro and Paz, 1993;
Kavlock et al.. 1980).
Table 7-7 provides a brief overview of the epidemiologic studies of fetal growth
restriction. In summary, the evidence is inconsistent when examining exposure to O3
and fetal growth restriction. Similar to PTB, studies conducted in Australia have
reported an effect of O3 on fetal growth, whereas studies conducted in other areas
generally have not found such an effect. This may be due to the restriction of births
to those within 2-14 km of a monitoring station, as was done in the Australian
studies.
7-56
-------
Table 7-7 Brief summary of epidemiologic studies of fetal growth.
Study
Salam et al.
(2005)
Braueret al.
(2008)
Liu et al. (2007b)
Hansen et al.
(2007)
Hansen et al.
(2008)
Manneset al.
(2005)
Location
(Sample Size)
California, U.S.
(n = 3,901)
Vancouver, BC, Canada
(n = 70,249)
Calgary, Edmonton, and
Montreal, Canada
(n = 1 6,430)
Brisbane, Australia
(n = 26,61 7)
Brisbane, Australia
(n = 15,623)
Sydney, Australia
(n = 138,056)
Mean O3 (ppb)
24-h avg:
27.3
8h:
50.6
24-h avg:
14
24-h avg:
16.5
1-h max:
31.2
8-h max:
26.7
8-h avg:
24.8
1-h max:
31.6
Exposure
assessment
ZIP code level
Nearest Monitor
(within 10 km)
Inverse Distance
Weighting (IDW)
Census Subdivision
avg
Citywide avg
Within 2 km of monitor
Citywide avg and
<5 km from monitor
Effect Estimate3 (95% Cl)
Entire pregnancy: 1.16 (1.00, 1.32)
11:1.00(0.94, 1.11)
T2: 1.06(1.00, 1.12)
T3: 1.11 (1.00,1.17)
Entire pregnancy: NR
First 30 days of pregnancy: NR
Last 30 days of pregnancy: NR
T1: NR
T3: NR
Entire pregnancy: NR (results
presented in figure)
T1 : NR (results presented in figure)
T2: NR (results presented in figure)
T3: NR (results presented in figure)
T1: 1.01 (0.89, 1.15)
T2: 1.00(0.86,1.17)
T3: 0.83 (0.71 , 0.97)
M1 : -0.32 (-1.56, 0.91 )b
M2: -0.58 (-1.97,0.80)"
M3: 0.26 (-1.07, 1 .59)b
M4:0.11 (-0.98, 1.21)"
T1 : 0.90 (0.48, 1 .34)
T2: 1 .00 (0.60, 1 .79)
T3: 1.10(0.66, 1.97)
Last 30 days of pregnancy: 1.10 (0.74,
1.79)
"Relative risk of fetal growth restriction per 10 ppb change in O3, unless otherwise noted.
bMean change in fetal ultrasonic measure of head circumference recorded between 13 and 26 weeks gestation for a 10-ppb
increase in maternal exposure to O3 during early pregnancy
T1 = First Trimester, T2 = Second Trimester, T3 = Third Trimester
M1 = Month 1, M2 = Month 2, M3 = Month 3, M4 = Month 4
NR: No quantitative results reported
7.4.6 Postnatal Growth
Postnatal weight and height are routinely measured in children as indicators of
growth and somatic changes. Toxicological studies often follow these endpoints to
ascertain if a known exposure has an effect in the postnatal window, an effect which
can be permanent. Time-pregnant BALB/c mice were exposed to O3 (0, 0.4, 0.8, or
1.2 ppm) GD9-18 (short-term exposure) with parturition at GD20-21 (Sharkhuu et
al., 2011). As the offspring aged, postnatal litter body weight continued to be
significantly decreased in the highest concentration (1.2 ppm) O3 group at PND3 and
PND7. When the pups were weighed separately by sex at PND42, the males with the
7-57
-------
highest concentration of O3 exposure (1.2 ppm, GD9-18) had significant decrements
in body weight (Sharkhuu et al., 2011).
Significant decrements in body weight at 4 weeks of age were reported in C57B1/6
mice that were exposed to postnatal O3 (short-term exposure, PND2-28 exposure,
1 ppm O3, 3 hours/day, 3 days/week) (Auten et al.. 2012). Animals with co-exposure
to in utero DE (short-term exposure, dam GD9-GD17; inhalation 0.5 or 2.0 mg/m3
O3; 4 h/day via inhalation; or oropharyngeal aspiration DEPs, 2x/week) + postnatal
O3 (aforementioned short-term exposure) also had significantly reduced body
weight.
7.4.7 Birth Defects
Despite the growing body of literature evaluating the association between ambient air
pollution and various adverse birth outcomes, relatively few studies have
investigated the effect of temporal variations in ambient air pollution on birth
defects. Heart defects and oral clefts have been the focus of the majority of these
recent studies, given the higher prevalence than other birth defects and associated
mortality. Mechanistically, air pollutants could be involved in the etiology of birth
defects via a number of key events (see Section 5.4.2.4).
Several studies have been conducted examining the relationship between O3
exposure during pregnancy and birth defects and reported a positive association with
cardiac defects. The earliest of these studies was conducted in southern California
(Ritz et al., 2002). This study evaluated the effect of air pollution on the occurrence
of cardiac birth defects in neonates and fetuses delivered in southern California in
1987-1993. Maternal exposure estimates were based on data from the fixed site
closest to the mother's ZIP code area. When using a case-control design where cases
were matched to 10 randomly selected controls, results showed increased risks for
aortic artery and valve defects (OR = 1.56 [95% CI: 1.16, 2.09] per 10 ppb O3),
pulmonary artery and valve anomalies (OR = 1.34 [95% CI: 0.96, 1.87] per 10 ppb
O3), and conotruncal defects (OR = 1.36 [95% CI: 0.91, 2.03] per 10 ppb O3) in a
dose-response manner with second-month O3 exposure. A study conducted in Texas
(Gilboa et al.. 2005) looked at a similar period of exposure but reported no
association with most of the birth defects studied (O3 concentration was studied
using quartiles with the lowest representing <18 ppb and the highest representing >
31 ppb). The authors found slightly elevated odds ratios for pulmonary artery and
valve defects. They also detected an inverse association between O3 exposure and
isolated ventricular septal defects. Overall, this study provided some weak evidence
that air pollution increases the risk of cardiac defects. Hansen et al. (2009)
investigated the possible association between ambient air pollution concentrations
averaged over weeks 3-8 of pregnancy and the risk of cardiac defects. When
analyzing all births with exposure estimates for O3 from the nearest monitor there
was no indication for an association with cardiac defects. There was also no adverse
association when restricting the analyses to only include births where the mother
7-58
-------
resided within 12 km of a monitoring station. However, among births within 6 km of
a monitor, a 10 ppb increase in O3 was associated with an increased risk of
pulmonary artery and valve defects (OR = 8.76 [95% CI: 1.80, 56.55]). As indicated
by the very wide credible intervals, there were very few cases in the sensitivity
analyses for births within 6 km of a monitor, and this effect could be a result of type I
errors. Dadvand et al. (2011) investigated the association between maternal exposure
to ambient air pollution concentrations averaged over weeks 3-8 of pregnancy and
the occurrence of cardiac birth defects in England. Similar to Hansen et al. (2009).
they found no associations with maternal exposure to O3 except for when the
analysis was limited to those subjects residing within a 16 km distance of a
monitoring station (OR for malformations of pulmonary and tricuspid valves=1.64
[95% CI: 1.04, 2.60] per 10 ppb increase in O3).
Despite the association between O3 and cardiac defects observed in the above
studies, a recent study did not observe an increased risk of cardiac birth defects
associated with ambient O3 concentrations. The study, conducted in Atlanta, GA,
examined O3 exposure during weeks 3-7 of of pregnancy and reported no association
with risk of cardiovascular malformations (Strickland et al., 2009).
Several of these studies have also examined the relationship between O3 exposure
during pregnancy and oral cleft defects. The study by Ritz et al. (2002) evaluated the
effect of air pollution on the occurrence of orofacial birth defects and did not observe
strong associations between ambient O3 concentration and orofacial defects. They
did report an OR of 1.13 (95% CI: 0.90, 1.40) per 10 ppb during the second trimester
for cleft lip with or without cleft palate. Similarly, Gilboa et al. (2005) reported an
OR of 1.09 (95% CI: 0.70, 1.69) for oral cleft defects when the fourth quartile was
contrasted with the first quartile of exposure during 3-8 weeks of pregnancy. Hansen
et al. (2009) reported no indication for an association with cleft defects and air
pollution concentrations averaged over weeks 3-8 of pregnancy. Hwang and Jaakkola
(2008) conducted a population-based case-control study to investigate exposure to
ambient air pollution and the risk of cleft lip with or without cleft palate in Taiwan.
The risk of cleft lip with or without cleft palate was increased in relation to O3 levels
in the first gestational month (OR =1.17 [95% CI: 1.01, 1.36] per 10 ppb) and
second gestational month (OR = 1.22 [95% CI: 1.03, 1.46] per 10 ppb), but was not
related to any of the other pollutants. In three-pollutant models, the effect estimates
for O3 exposure were stable for the four different combinations of pollutants and
were all statistically significant. Marshall et al. (2010) compared estimated exposure
to ambient pollutants during early pregnancy (6 week period from 5 to 10 weeks into
the gestational period) among mothers of children with oral cleft defects to that
among mothers of controls. The authors observed no consistent elevated associations
between any of the air pollutants examined and cleft malformations, though there
was a weak association between cases of cleft palate only and increasing O3
concentrations. This association increased when cases and controls were limited to
those with residences within 10 km of the closest O3 monitor (OR = 2.2 [95% CI:
1.0, 4.9], comparing highest quartile [>33 ppb] to lowest quartile [<15 ppb]).
7-59
-------
A limited number of toxicological studies have examined birth defects in animals
exposed gestationally to O3. Kavlock et al. (1979) exposed pregnant rats to O3 for
precise periods during organogenesis. No significant teratogenic effects were found
in rats exposed 8 h/day to concentrations of O3 varying from 0.44 to 1.97 ppm during
early (days 6-9), mid (days 9-12), or late (days 17 to 20) gestation, or the entire
period of organogenesis (days 6-15) (short-term exposures). Earlier research found
eyelid malformation following gestational and postnatal exposure to 0.2 ppm O3
(Veninga. 1967).
Table 7-8 provides a brief overview of the epidemiologic studies of birth defects.
These studies have focused on cardiac and oral cleft defects, and the results from
these studies are not entirely consistent. This inconsistency could be due to the
absence of true associations between O3 and risks of cardiovascular malformations
and oral cleft defects; it could also be due to differences in populations, pollution
levels, outcome definitions, or analytical approaches. The lack of consistency of
associations between O3 and cardiovascular malformations or oral cleft defects might
be due to issues relating to statistical power or measurement error. A recent meta-
analysis of air pollution and congenital anomalies concluded that there was no
statistically significant increase in risk of congenital anomalies with O3 exposure
(Vrijheid et al., 2011). These authors note that heterogeneity in the results of these
studies may be due to inherent differences in study location, study design, and/or
analytic methods, and comment that these studies have not employed some recent
advances in exposure assessment used in other areas of air pollution research that
may help refine or reduce this heterogeneity.
7-60
-------
Table 7-8 Brief summary of epidemiologic studies of birth defects.
Study
Ritz et al. (2002)
Gilboa et al. (2005)
Hwang and Jaakkola
(2008)
Strickland et al.
(2009)
Hansen et al. (2009)
Marshall et al. (2010)
Dadvand et al.
(2011)
Outcomes
Examined
Cardiac and Cleft
Defects
Cardiac and Cleft
Defects
Oral Cleft Defects
Cardiac Defects
Cardiac and Cleft
Defects
Oral Cleft Defects
Cardiac Defects
Location
(Sample Size)
Southern California
(n = 3,549 cases;
10,649 controls)
7 Counties in TX
(n = 5,338 cases;
4,580 controls)
Taiwan
(n = 653 cases;
6,530 controls)
Atlanta, GA
(n = 3,338 cases)
Brisbane, Australia
(n = 150,308 births)
New Jersey
(n = 717 cases;
12,925 controls)
Northeast England
(n = 2, 140 cases;
14,256 controls)
Mean Os (ppb)
24-h avg:
NR
24-h avg:
NR
24-h avg:
27.31
8-h max:
39.8-43.3
8-h max:
25.8
24-h avg:
25
24-h avg:
18.8
Exposure
Assessment
Nearest Monitor
(within 10 mi)
Nearest Monitor
Inverse Distance
Weighting (IDW)
Weighted citywide
avg
Nearest Monitor
Nearest Monitor
(within 40 km)
Nearest Monitor
Exposure
Window
Month 1,2,3
Trimester 2,3
3-mo period prior to
conception
Weeks 3-8 of
gestation
Months 1,2,3
Weeks 3-7 of
gestation
Weeks 3-8 of
gestation
Weeks 5-1 Oof
gestation
Weeks 3-8 of
gestation" 1
7.4.8 Developmental Respiratory Effects
The issue of prenatal exposure has assumed increasing importance since ambient air
pollution exposures of pregnant women have been shown to lead to adverse
pregnancy outcomes, as well as to respiratory morbidity and mortality in the first
year of life. Growth and development of the respiratory system take place mainly
during the prenatal and early postnatal periods. This early developmental phase is
thought to be very important in determining long-term lung growth. Studies have
recently examined this emerging issue. Several studies were included in Section 7.2.1
and Section 7.2.3. and are included here because they reported both prenatal and
post-natal exposure periods.
Mortimer et al. (2008a, b) examined the association of prenatal and lifetime
exposures to air pollutants with pulmonary function and allergen sensitization in a
subset of asthmatic children (ages 6-11) included in the Fresno Asthmatic Children's
Environment Study (FACES). Monthly means of pollutant levels for the years
1989-2000 were created and averaged separately across several important
developmental time-periods, including the entire pregnancy, each trimester, the first
3 years of life, the first 6 years of life, and the entire lifetime. The 8-h avg O3
concentrations were approximately 50 ppb for each of the exposure metrics
7-61
-------
(estimated from figure). In the first analysis (Mortimer et al., 2008a), negative effects
on pulmonary function were found for exposure to PMi0, NO2, and CO during key
neonatal and early life developmental periods. The authors did not find a negative
effect of exposure to O3 among this cohort. In the second analysis (Mortimer et al.,
2008b), sensitization to at least one allergen was associated, in general, with higher
levels of CO and PMi0 during the entire pregnancy and second trimester and higher
PMio during the first 2 years of life. Lower exposure to O3 during the entire
pregnancy or second trimester was associated with an increased risk of allergen
sensitization. Although the pollutant metrics across time periods are correlated, the
strongest associations with the outcomes were observed for prenatal exposures.
Though it may be difficult to disentangle the effect of prenatal and postnatal
exposures, the models from this group of studies suggest that each time period of
exposure may contribute independently to different dimensions of school-aged
children's pulmonary function. For 4 of the 8 pulmonary-function measures (FVC,
FEVi, PEF, FEF25.75), prenatal exposures were more influential on pulmonary
function than early-lifetime metrics, while, in contrast, the ratio of measures
(FEVi/FVC and FEF25-75/FVC) were most influenced by postnatal exposures. When
lifetime metrics were considered alone, or in combination with the prenatal metrics,
the lifetime measures were not associated with any of the outcomes, suggesting the
timing of the exposure may be more important than the overall dose and prenatal
exposures are not just markers for lifetime or current exposures.
Clark et al. (2010) investigated the effect of exposure to ambient air pollution in
utero and during the first year of life on risk of subsequent asthma diagnosis (incident
asthma diagnosis up to age 3-4) in a population-based nested case-control study. Air
pollution exposure for each subject based on their residential address history was
estimated using regulatory monitoring data, land use regression modeling, and
proximity to stationary pollution sources. An average exposure was calculated for the
duration of pregnancy (~15 ppb) and the first year of life (~14 ppb). In contrast to the
Mortimer et al. (2008a, b) studies, the effect estimates for first-year exposure were
generally larger than for in utero exposures. However, similar to the Mortimer et al.
(2008a, b), the observed associations with O3 were largely protective. Because of the
relatively high correlation between in utero and first-year exposures for many
pollutants, it was difficult to discern the relative importance of the individual
exposure periods.
Latzin et al. (2009) examined whether prenatal exposure to air pollution was
associated with lung function changes in the newborn. Tidal breathing, lung volume,
ventilation inhomogeneity and eNO were measured in 241 unsedated, sleeping
neonates (age = 5 weeks). The median of the 24-h avg O3 concentrations averaged
across the post-natal period was -44 ppb. Consistent with the previous studies, no
association was found for prenatal exposure to O3 and lung function.
The new toxicological literature since the 2006 O3 AQCD, covering respiratory
changes related to developmental O3 exposure, reports ultrastructural changes in
bronchiole development, alterations in placental and pup cytokines, and increased
pup airway hyper-reactivity. These studies are detailed below.
7-62
-------
Fetal rat lung bronchiole development is triphasic, comprised of the glandular phase
(measured at GDIS), the canalicular phase (GD20), and the saccular phase (GD21).
The ultrastructural lung development in fetuses of pregnant rats exposed to 1-ppm O3
(12 h/day, out to either GDIS, GD20 or GD21) was examined by electron
microscopy during these three phases. In the glandular phase, bronchiolar columnar
epithelial cells in fetuses of dams exposed to O3 had cytoplasmic damage and
swollen mitochondria. Bronchial epithelium at the canalicular phase in O3 exposed
pups had delayed maturation in differentiation, i.e., glycogen abundance in secretory
cells had not diminished as it should with this phase of development. Congruent with
this finding, delayed maturation of tracheal epithelium following early neonatal O3
exposure (1 ppm, 4-5 h/day for first week of life) in lambs has been previously
reported (Mariassv et al.. 1990: Mariassv et al.. 1989). Also at the canalicular phase,
atypical cells were seen in the bronchiolar lumen of O3-exposed rat fetuses. Finally,
in the saccular phase, mitochondrial degradation was present in the non-ciliated
bronchiolar cells of rats exposed in utero to O3. In conclusion, O3 exposure of
pregnant rats produced ultra-structural damage to near-term fetal bronchiolar
epithelium (Lopez et al.. 2008).
Exposure of laboratory animals to multiple airborne pollutants can differentially
affect pup physiology. One study showed that exposure of C57BL/6 mouse dams to
0.48 mg PM intratracheally twice weekly for 3 weeks during pregnancy augmented
O3-induced airway hyper-reactivity in juvenile offspring. Maternal PM exposure also
significantly increased placental cytokines above vehicle-instilled controls. Pup
postnatal O3 exposure (1 ppm 3 h/day, every other day, thrice weekly for 4 weeks)
induced significantly increased cytokine levels (IL-1(3, TNF-a, KC, and IL-6) in
whole lung versus postnatal air exposed groups; this was further exacerbated with
gestational PM exposure (Auten et al.. 2009). In further studies by the same
laboratory, O3-induced AHR was studied in rodent offspring after dam gestational
exposure to inhaled diesel exhaust (Auten et al.. 2012). Pregnant C57B1/6 mice were
exposed to diesel exhaust GD9-17 (0.5 or 2.0 mg/m3 O3, 4h/day) via inhalation or in
a separate set of animals via oropharyngeal aspiration of freshly generated DEPs
(2x/week). Postnatally, the offspring were exposed to O3 starting at PND2 (1 ppm
O3, 3 hours/day, 3 days/week for 4 weeks). Juvenile mice were then subjected to
measurements of pulmonary mechanisms (at 4 weeks of age and then at 8 weeks of
age). Increased inflammation of the placenta and lungs of DE exposed fetuses was
reported at GDI8. In animals with postnatal O3 exposure alone, elevated
inflammation was seen with significant increased levels of BAL cytokines; these O3-
related elevated levels were significantly exacerbated with prenatal DE exposure
(DE+O3). At PND28, DE+O3 exposed offspring had significant impairment of
alveolar development as measured with secondary alveolar crest development, a
finding that was absent in all other exposure groups (O3 alone, DE alone). Postnatal
O3 exposure induced AHR in methacholine challenged animals at 4 weeks of age and
was exacerbated with the higher dose of DE exposure (DE+O3). At 8 weeks of age,
O3 exposed pups had persistent AHR (+/-DE) that was significantly augmented in
DE+O3 pups. In summary, gestational DE exposure induced an inflammatory
response which, when combined with postnatal O3 exposure impaired alveolar
7-63
-------
development, and caused an exacerbated and longer-lasting O3-induced AHR in
offspring.
A series of experiments using infant rhesus monkeys repeatedly exposed to 0.5 ppm
O3 starting at one-month of age have examined the effect of O3 alone or in
combination with an inhaled allergen on morphology and lung function (Plopper et
al., 2007). Exposure to O3 alone or allergen alone produced small but not statistically
significant changes in baseline airway resistance and airway responsiveness, but the
combined exposure to both O3 + antigen produced statistically significant and greater
than additive changes in both functional measurements. Additionally, cellular
changes and significant structural changes in the respiratory tract have been observed
in infant rhesus monkeys exposed to O3 (Fanucchi et al., 2006). A more detailed
description of these studies can be found in Section 7.2.3 (Pulmonary Structure and
Function), with mechanistic information found in Section 5.4.2.4.
Lung immunological response in O3 exposed pups was followed by analyzing BAL
and lung tissue. Sprague Dawley (SD) pups were exposed to a single 3h exposure of
air or O3 (0.6 ppm) on PND 13 (Han et al.. 2011). Bronchoalveolar lavage (BAL)
was performed 10 hours after the end of O3 exposure. BALF polymorphonuclear
leukocytes (PMNs) and total BALF protein were significantly elevated in O3
exposed pups. Lung tissue from O3 exposed pups had significant elevations of
manganese superoxide dismutase (SOD) protein and significant decrements of extra-
cellular SOD protein.
Various immunological outcomes were followed in offspring after their pregnant
dams (BALB/c mice) were exposed gestationally to O3 (0, 0.4, 0.8, or 1.2 ppm,
GD9-18) (Sharkhuu et al., 2011). Delayed type hypersensitivity (DTH) was initiated
with initial BSA injection at 6 weeks of age and then challenge 7 days later.
The normal edematous response of the exposed footpad (thickness after BSA
injection) was recorded as an indicator of DTH. In female offspring, normal footpad
swelling with BSA injection that was seen in air exposed animals was significantly
attenuated with O3 exposure (0.8 and 1.2 ppm O3), implying immune suppression of
O3 exposure specifically in DTH. Humoral immunity was measured with the sheep
red blood cell (SRBC) response. Animals received primary immunization with
SRBC and then blood was drawn for SRBC IgM measurement. A SRBC booster was
given 2 weeks later with blood collected 5 days after booster for IgG measurement.
Maternal O3 exposure had no effect on humoral immunity in the offspring as
measured by IgG and IgM titers after SRBC primary and booster immunizations
(Sharkhuu et al.. 2011).
Toxicity assessment and allergen sensitization was also assessed in these O3 exposed
offspring. At PND42, animals were euthanized for analysis of immune and
inflammatory markers (immune proteins, inflammatory cells, T-cell populations in
the spleen). A subset of the animals was intra-nasally instilled or sensitized with
ovalbumin on either PND2 and 3 or PND42 and 43. All animals were challenged
with OVA on PND54, 55, and 56. One day after final OVA challenge, lung function,
lung inflammation and immune response were determined. Offspring of O3 exposed
dams that were initially sensitized at PDN3 (early) or PND42 (late) were tested to
7-64
-------
determine the level of allergic sensitization or asthma-like inflammation after OVA
challenge. Female offspring sensitized early in life developed significant eosinophilia
(1.2 ppm O3) and elevated serum OVA-specific IgE (1.2 ppm O3), which is a marker
of airway allergic inflammation. The females that were sensitized early also had
significant decrements in BALF total cells, macrophages, and lymphocytes (1.2 ppm
O3). Offspring that were sensitized later (PND42) in life did not develop the
aforementioned changes in BALF, but these animals did develop modest, albeit
significant neutropenia (0.8 and 1.2 ppm O3) (Sharkhuu et al.. 2011).
BALF cytology in non-sensitized animals was followed. BALF of offspring born to
dams exposed to O3 was relatively unaffected (cytokines, inflammatory cell
numbers/types) as were splenic T-cell subpopulations. LDH was significantly
elevated in BALF of females whose mothers were exposed to 1.2 ppm during
pregnancy (Sharkhuu et al., 2011). In summary, the females born to mothers exposed
to O3 developed modest immunocompromise. Males were unaffected (Sharkhuu et
al..20m
Overall, animal toxicological studies have reported ultrastructural changes in
bronchiole development, alterations in placental and pup cytokines, and increased
pup airway hyper-reactivity related to exposure to O3 during the developmental
period. Epidemiologic studies have found no association between prenatal exposure
to O3 and growth and development of the respiratory system. Fetal origins of disease
have received a lot of attention recently, thus additional research to further explore
the inconsistencies between these two lines of evidence is warranted.
7.4.9 Developmental Central Nervous System Effects
The following sections describe the results of toxicological studies of O3 and
developmental central nervous system effects. No epidemiologic studies of this
association have been published.
7.4.9.1 Laterality
Two reports of laterality changes in mice developmentally exposed to O3 have been
reported in the literature. Mice developmentally exposed to 0.6 ppm O3 (6 days
before breeding to weaning at PND21) showed a turning preference (left turns)
distinct from air exposed controls (clockwise turns) (Dell'Omo et al., 1995); in
previous studies this behavior in mice has been found to correlate with specific
structural asymmetries of the hippocampal mossy fiber projections (Schopke et al.,
1991). The 2006 O3 AQCD evidence for the effect of O3 on laterality or handedness
demonstrated that rats exposed to O3 during fetal and neonatal life showed limited,
sex-specific changes in handedness after exposure to the intermediate concentration
of O3 (only seen in female mice exposed to 0.6 ppm O3, and not in males at 0.6 ppm
7-65
-------
or in either sex of 0.3 or 0.9 ppm O3 with exposure from 6 days before breeding to
PND26) (Petruzzi et al.. 1999).
7.4.9.2 Brain Morphology and Neurochemical Changes
The nucleus tractus solitarius (NTS), a medullary area of respiratory control, of adult
animals exposed prenatally to 0.5 ppm O3 (12h/day, ED5-eD20) had significantly
less tyrosine hydroxylase staining versus control (Boussouar et al., 2009). Tyrosine
hydroxylase is the rate-limiting enzyme for dopamine synthesis and serves as a
precursor for catecholamine synthesis; thus, decreased staining is used as a marker of
dopaminergic or catecholaminergic cell or activity loss in these regions and thus
functions in neuronal plasticity. After physical restraint stress, control animals
respond at the histological level with Fos activation, a marker of neuronal activity,
and tyrosine hydroxylase activation in the NTS, a response which is absent or
attenuated in adult animals exposed prenatally to 0.5 ppm O3 (Boussouar et al.,
2009) when compared to control air exposed animals who also were restrained.
The O3-exposed offspring in this study were cross-fostered to control air exposed
dams to avoid O3-dependent dam related neonatal effects on offspring outcomes
(i.e., dam behavioral or lactational contributions to pup outcomes) (Boussouar et al.,
2009).
Developmental exposure to 0.3 or 0.6 ppm O3 prior to mating pair formation through
GDI 7 induced significant increased levels of BDNF in the striatum of adult
(PND140) O3-exposed offspring as compared to control air exposed animals; these
O3-exposed animals also had significantly decreased level of NGF in the
hippocampus versus control (Santucci et al.. 2006).
Changes in the pup cerebellum with prenatal 1 ppm O3 exposure include altered
morphology (Romero-Velazquez et al.. 2002: Rivas-Manzano andPaz. 1999).
decreased total area (Romero-Velazquez et al.. 2002). decreased number of Purkinje
cells (Romero-Velazquez et al.. 2002). and altered monoamine neurotransmitter
content with the catecholamine system affected and the indoleamine system
unaffected by O3 (Gonzalez-Pina et al.. 2008).
7.4.9.3 Neurobehavioral Outcomes
Ozone administration to dams during pregnancy with or without early neonatal
exposure has been shown to contribute to multiple neurobehavioral outcomes in
offspring that are described in further detail below.
Ozone administration (0.4, 0.8 or 1.2 ppm O3) during the majority of pregnancy
(PD7-17) of CD-I mice did not affect pup behavioral outcomes including early
behavioral ultrasonic vocalizations and more permanent later measurements (PND60
or 61) including pup activity, habituation and exploration and d-
7-66
-------
amphetamine-induced hyperactivity (Bignami et al., 1994); these pups were all cross-
fostered or reared on non- O3 exposed dams.
Testing for aggressive behavior in mice continuously exposed to O3 (0.3 or 0.6 ppm
from 30 days prior to mating to GD17) revealed that mice had significantly increased
defensive/ submissive behavior (increased freezing posturing on the first day only of
a multiple-day exam) versus air exposed controls (Santucci et al., 2006). Similarly,
continuous exposure of adult animals to O3 induced significant increases in fear
behavior and decreased aggression as measured by significantly decreased freezing
behavior (Petruzzi et al., 1995).
Developmentally exposed animals also had significantly decreased amount of time
spent nose sniffing other mice (Santucci et al.. 2006): this social behavior deficit,
decreased sniffing time, was not found in an earlier study with similar exposures
(Petruzzi et al.. 1995). but sniffing of specific body areas was measured in Santucci
et al. (2006) and total number of sniffs of the entire body was measured in Petruzzi et
al. (1995). The two toxicology studies exploring social behavior (sniffing) employ
different study designs and find opposite effects in animals exposed to O3.
7.4.9.4 Sleep Aberrations after Developmental Ozone Exposure
The effect of gestational O3 exposure (1 ppm O3 daily for 12h/day, during dark
period for the entire pregnancy) on sleep patterns in rat offspring was followed using
24 h polysomnographic recordings at 30, 60 and 90 days of age (Haro and Paz,
1993). Ozone-exposed pups manifested with inverted sleep-wake patterns or
circadian rhythm phase-shift. Rat vigilance was characterized in wakefulness, slow
wave sleep (SWS), and paradoxical sleep (PS) using previously characterized
criteria. The O3 exposed offspring spent longer time in the wakefulness state during
the light period, more time in SWS during the period of darkness, and showed
significant decrements in PS. Chronic O3 inhalation significantly decreased the
duration of PS during both the light and dark periods (Haro and Paz. 1993). These
effects were consistent at all time periods measured (30, 60 and 90 days of age).
These sleep effects reported after developmental exposures expand upon the existing
literature on sleep aberrations in adult animals exposed to O3 [rodents: (Paz and
Huitron-Resendiz. 1996: Arito et al.. 1992): and cats: (Paz and Bazan-Perkins.
1992)1. A role for inhibition of cyclooxygenase-2 and the interleukins and
prostaglandins in the O3-dependent sleep changes potentially exists with evidence
from a publication on indomethacin pretreatment attenuating O3-induced sleep
aberrations in adult male animals (Rubio and Paz. 2003).
7.4.10 Early Life Mortality
Infants may be particularly at risk for the effects of air pollution. Within the first year
of life, infants develop rapidly; therefore their sensitivity may change within weeks
7-67
-------
or months. During the neonatal and post-neonatal periods, the developing lung is
highly sensitive to environmental toxicants. The lung is not well developed at birth,
with 80% of alveoli being formed postnatally. An important question regarding the
association between O3 and infant mortality is the critical window of exposure
during development for which infants are at risk. Several age intervals have been
explored: neonatal (<1 month); postneonatal (1 month to 1 year); and an overall
interval for infants that includes both the neonatal and postneonatal periods
(<1 year). Within these various age categories, multiple causes of deaths have been
investigated, particularly total deaths and respiratory-related deaths. The studies
reflect a variety of study designs, exposure periods, regions, and adjustment for
confounders. As discussed below, a handful of studies have examined the effect of
ambient air pollution on neonatal and postneonatal mortality, with the former the
least studied. These studies varied somewhat with regard to the outcomes and
exposure periods examined and study designs employed.
7.4.10.1 Stillbirth
Pereira et al. (1998) investigated the association among daily counts of intrauterine
mortality (over 28 weeks of gestation) and air pollutant concentrations in Sao Paulo,
Brazil from 1991 through 1992. The association was strong for NO2, but lesser for
SO2 and CO. These associations exhibited a short lag time, less than 5 days.
No significant association was detected between short-term O3 exposure and
intrauterine mortality.
7.4.10.2 Infant Mortality, Less than 1 Year
Ritz et al. (2006) linked birth and death certificates for infants who died between
1989 and 2000 to evaluate the influence of outdoor air pollution on infant death in
the South Coast Air Basin of California. The authors examined short- and long-term
exposure periods 2 weeks, 1 month, 2 months, and 6 months before each case
subject's death and reported no association between ambient levels of O3 and infant
mortality. Similarly, Diaz et al. (2004) analyzed the effects of extreme temperatures
and short-term exposure to air pollutants on daily mortality in children less than
1 year of age in Madrid, Spain, from 1986 to 1997 and observed no statistically
significant association between mortality and O3 concentrations. Hajat et al. (2007)
analyzed time-series data of daily infant mortality counts in 10 major cities in the UK
to quantify any associations with short-term changes in air pollution. When the
results from the 10 cities were combined there was no relationship between O3 and
infant mortality, even after restricting the analysis to just the summer months.
Conversely, a time-series study of infant mortality conducted in the southwestern
part of Mexico City in the years 1993-1995 found that infant mortality was
associated with short-term exposure to NO2 and O3 3-5 days before death, but not as
consistently as with PM. A 10-ppb increase in 24-h avg O3 was associated with a
7-68
-------
2.78% increase (95% CI: 0.29, 5.26%) in infant mortality (lag 3) (Loomis et al..
1999). This increase was attenuated, although still positive when evaluated in a two-
pollutant model with PM2.5. One-hour max concentrations of O3 exceeded prevailing
Mexican and international standards nearly every day.
7.4.10.3 Neonatal Mortality, Less than 1 Month
Several studies have evaluated ambient O3 concentrations and neonatal mortality and
observed no association. Ritz et al. (2006) linked birth and death certificates for
infants who died between 1989 and 2000 to evaluate the influence of outdoor air
pollution on infant death in the South Coast Air Basin of California. The authors
examined short- and long-term exposure periods 2 weeks, 1 month, 2 months, and
6 months before each case subject's death and reported no association between
ambient levels of O3 and neonatal mortality. Hajat et al. (2007) analyzed time-series
data of daily infant mortality counts in 10 major cities in the UK to quantify any
associations with short-term changes in air pollution. When the results from the 10
cities were combined there was no relationship between O3 and neonatal mortality,
even after restricting the analysis to just the summer months. Lin et al. (2004a)
assessed the impact of short-term changes in air pollutants on the number of daily
neonatal deaths in Sao Paulo, Brazil. The authors observed no association between
ambient levels of O3 and neonatal mortality.
7.4.10.4 Postneonatal Mortality, 1 Month to 1 Year
A number of studies focused on the postneonatal period when examining the effects
of O3 on infant mortality. Ritz et al. (2006) linked birth and death certificates for
infants who died between 1989 and 2000 to evaluate the influence of outdoor air
pollution on infant death in the South Coast Air Basin of California. The authors
examined short- and long-term exposure periods 2 weeks, 1 month, 2 months, and
6 months before each case subject's death and reported no association between
ambient levels of O3 and postneonatal mortality. Woodruff et al. (2008) evaluated
the county-level relationship between cause-specific postneonatal infant mortality
and long-term early-life exposure (first 2 months of life) to air pollutants across the
United States. Similarly, they found no association between O3 exposure and deaths
from respiratory causes. In the U.K., Hajat et al. (2007) analyzed time-series data of
daily infant mortality counts in 10 major cities to quantify any associations with
short-term changes in air pollution. When the results from the 10 cities were
combined there was no relationship between O3 and postneonatal mortality, even
after restricting the analysis to just the summer months. In Ciudad Juarez, Mexico,
Romieu et al. (2004a) examined the daily number of deaths between 1997 and 2001,
estimating the modifying effect of SES on the risk of postneonatal mortality.
Ambient O3 concentrations were not related to infant mortality overall, or in any of
the SES groups. In a follow-up study, Carbajal-Arroyo et al. (2011) evaluated the
7-69
-------
relationship of 1-h daily max O3 levels with postneonatal infant mortality in the
Mexico City Metropolitan Area between 1997 and 2005. Generally, short-term
exposure to O3 was not significantly related to infant mortality. However, upon
estimating the modifying effect of SES on the risk of postneonatal mortality, the
authors found that O3 was statistically significantly related to respiratory mortality
among those with low SES. In a separate analysis, the effect of PMi0 was evaluated
with O3 level quartiles. PMi0 alone was related to a significant increase in all-cause
mortality. The magnitude of this effect remained the same when only the days when
O3 was in the lowest quartile were included in the analyses. However, when only the
days when O3 was in the highest quartile were included in the analyses, the
magnitude of the PM10 effect increased dramatically (OR = 1.06 [95% CI: 0.909,
1.241] for PM10 on days with O3 in lowest quartile; OR = 1.26 [95% CI: 1.08, 1.47]
for PMio on days with O3 in the highest quartile. These results suggest that while O3
alone may not have an effect on infant mortality, it may serve to potentiate the
observed effect of PMi0 on infant mortality.
Tsai et al. (2006) used a case-crossover analysis to examine the relationship between
short-term exposure to air pollution and postneonatal mortality in Kaohsiung, Taiwan
during the period 1994-2000. The risk of postneonatal deaths was 1.023 (95% CI:
0.564, 1.858) per 10-ppb increase in 24-h avg O3. The confidence interval for this
effect estimate is very wide, likely due to the small number of infants that died each
day, making it difficult to interpret this result. Several other studies conducted in
Asia did not find any association between O3 concentrations and infant mortality in
the postneonatal period. Ha et al. (2003) conducted a daily time-series study in Seoul,
Korea to evaluate the effect of short-term changes in ambient 8-h O3 concentrations
on postneonatal mortality. Son et al. (2008) examined the relationship between air
pollution and postneonatal mortality from all causes among firstborn infants in Seoul,
Korea during 1999-2003. Yang et al. (2006) used a case-crossover analysis to
examine the relationship between air pollution exposure and postneonatal mortality
in Taipei, Taiwan for the period 1994-2000. The authors observed no associations
between ambient levels of O3 and postneonatal mortality.
7.4.10.5 Sudden Infant Death Syndrome
The strongest evidence for an association between ambient O3 concentrations and
SIDS comes from a study that evaluated the county-level relationship between SIDS
and long-term early-life exposure (first 2 months of life) to air pollutants across the
U.S. (Woodruff etal.. 2008). The authors observed a 1.20 (95% CI: 1.09, 1.32) odds
ratio for a 10-ppb increase in O3 and deaths from SIDS. There was a monotonic
increase in odds of SIDS for each quartile of O3 exposure compared with the lowest
quartile (highest quartile OR = 1.51; [95% CI: 1.17, 1.96]). In a multipollutant model
including PM10 or PM2.5, CO and SO2, the OR for SIDS and O3 was not
substantially lower than that found in the single-pollutant model. When examined by
season, the relationship between SIDS deaths and O3 was generally consistent across
seasons with a slight increase for those babies born in the summer. When stratified
7-70
-------
by birth weight, the OR for LEW babies was 1.27 (95% CI: 0.95, 1.69) per 10-ppb
increase in O3 and the OR for normal weight babies was 1.16 (95% CI: 1.01, 1.32)
per 10-ppb increase in O3.
Conversely, two additional studies reported no association between ambient levels of
O3 and SIDS. Ritz et al. (2006) linked birth and death certificates for infants who
died between 1989 and 2000 to evaluate the influence of outdoor air pollution on
infant death in the South Coast Air Basin of California. The authors examined short-
and long-term exposure periods 2 weeks, 1 month, 2 months, and 6 months before
each case subject's death and reported no association between ambient levels of O3
and SIDS. Dales et al. (2004) used time-series analyses to compare the daily
mortality rates for SIDS and short-term air pollution concentrations in 12 Canadian
cities during the period of 1984-1999. Increased daily rates of SIDS were associated
with previous day increases in the levels of SO2, NO2, and CO, but not O3 or PM2.5.
Table 7-9 provides a brief overview of the epidemiologic studies of infant mortality.
These studies have focused on short-term exposure windows (e.g., 1-3 days) and
long-term exposure windows (e.g., up to 6 months). Collectively, they provide no
evidence for an association between ambient O3 concentrations and infant mortality.
Table 7-9 Brief summary of infant mortality studies.
Study
Pereira et al.
(1998)
Diaz et al. (2004)
Loomis et al.
(1999)
Ritz et al. (2006)
Haiat et al. (2007)
Lin et al. (2004a)
Location
Sao Paulo,
Brazil
Madrid, Spain
Mexico City,
Mexico
Southern
California
10 Cities in the
UK
Sao Paulo,
Brazil
Mean O3 (ppb)
1-h max:
33.8
24-h avg:
11.4
24-h avg:
44.1
1-h max:
163.5
24-h avg:
21.9-22.1
24-h avg:
20.5-42.6
24-h avg:
38.06
Exposure
Assessment
City wide avg
City wide avg
1 monitor
Nearest Monitor
City wide avg
Citywide avg
Effect Estimate3 (95% CI):
LO-2: 1.00(0.99, 1.01)
NR
LO: 0.99(0.97, 1.02)
L1: 0.99(0.96, 1.01)
L2: 1.00(0.98, 1.03)
L3: 1.03(1.00, 1.05)
L4: 1.01 (0.98, 1.03)
L5: 1.02(0.99, 1.04)
LO-2: 1.02(0.99, 1.05)
2 weeks before death: 1.03 (0.93,
1.14)
1 mo before death: NR
2 mo before death: 0.93 (0.89,
0.97)
6 mo before death: NR
LO-2: 1.00(0.96, 1.06)
LO: 1.00(0.99, 1.01)
7-71
-------
Study
Ha et al. (2003)
Romieu et al.
(2004a)
Carbajal-Arroyo et
al. (2011)
Son et al. (2008)
Tsai et al. (2006)
Woodruff et al.
(2008)
Yang et al. (2006)
Dales et al. (2004)
Location
Seoul, South
Korea
Ciudad Juarez,
Mexico
Mexico City,
Mexico
Seoul, South
Korea
Kaohsiung,
Taiwan
Nationwide,
U.S.
Taipei, Taiwan
12 Canadian
Mean O3 (ppb)
8-h avg:
21.2
8-h avg:
43.43-55.12
1-h max:
103.0
8-h avg:
25.61
24-h avg:
23.60
24-h avg:
26.6
24-h avg:
18.14
24-h: 31.77
Exposure
Assessment
City wide avg
City wide avg
City wide avg
City wide avg
Citywide avg
County wide avg
Citywide avg
Citywide avg
Effect Estimate3 (95% Cl):
LO: 0.93(0.90, 0.96)
L1: 0.96(0.90, 1.03)
L2: 0.97(0.91, 1.04)
LO-1 cum: 0.96(0.89, 1.04)
L0-2cum: 0.94(0.87, 1.02)
LO: 1.00(0.99, 1.00)
L1: 0.99(0.99, 0.99)
L2: 0.99(0.99, 1.00)
LO-2: 0.99(0.99, 1.00)
L(NR): 0.984 (0.976, 0.992)b
LO-2 cum: 1.02(0.56, 1.86)
First 2 mo of life: 1.04(0.98, 1.10)
LO-2 cum:1. 00 (0.62, 1.61)
LO: NR
cities
L1: NR
L2: NR
L3: NR
L4: NR
L5: NR
Multiday lags of 2-6 days: NR
aRelative risk of infant mortality per 10 ppb change in O3
bNo increment provided
LO = Lag 0, L1 = Lag 1, L2 = Lag 2, L3 = Lag 3, L4 = Lag 4, L5 = Lag 5, L6 = Lag 6
NR: No quantitative results reported
7-72
-------
Table 7-10 Summary of key reproductive and developmental toxicological
studies.
Study
Sharkhuu et
al. (2011)
Biqnami et
al. (1994)
Haro and
Paz(1993)
Lopez et al.
(2008)
Auten et al.
(2009)
Plopper et
al. (2007)
Fanucchi et
al. (2006)
Dell'Omo et
al. (1995)
Santucci et
al. (2006)
03
Model (ppm)
Pregnant 0.4,
mice; BALB/c; 0.8, or
F;GD9-18; 1.2
effects in
offspring
Pregnant CD- 0.4,
1 dams (PD7- 0.8 or
17) 1.2
Rat dams, 1.0
Exposure over
the entirety of
pregnancy;
Rats; 1.0
Pregnant
dams; GDI-
GDIS, GD20,
orGD21.
C57BL/6 1.0
mouse pups
Infant rhesus 0.5
monkeys
Infant male 0.5
Rhesus
monkeys,
post-natal
exposure
CD-1 Mouse 0.6
dams and
pups
CD-1 Mouse 0.3 or
dams 0.6
Exposure Duration
Continuously for 10
consecutive days
Continuous
12h/day during dark
cycle
(12h/day, out to
either GD1 8, GD20
orGD21)
3 h/day, every other
day, thrice weekly
for 4 weeks
Postnatal, PND30-
6month of age, 5
months of cyclic
exposure, 5 days O3
followed by 9 days
of filtered air,
8h/day.
5 months of episodic
exposure, age 1
month-age 6
months, 5 days O3
followed by 9 days
of filtered air,
8 h/day.
6 days before
breeding to weaning
at PND21
Dam exposure prior
to mating through
Effects
Dams: Decreased number of dams reaching
parturition. Offspring: (l)-Decreased birth weights.
(2)-Decreased rate of postnatal growth (body
weight). (S)-impaired delayed type
hypersensitivity.(4)-No effect on humoral immunity.
(S)-Significantly affected allergic airway
inflammation markers (eosinophilia, IgE) in female
offspring sensitized early in life. 6-BALF LDH
significantly elevated in female offspring.
Reproductive success was not affected by O3
exposure (PD7-17, proportion of successful
pregnancies, litter size, ex ratio, frequency of still
birth, or neonatal mortality). Ozone acted as a
transient anorexigen in pregnant dams.
Decreased birth weight and postnatal body weight
of offspring out to PND 90. Ozone-exposed pups
manifested with inverted sleep-wake patterns or
circadian rhythm phase-shift.
O3 induced delayed maturation of near term
rodent bronchioles, with ultra-structural damage to
bronchiolar epithelium.
Postnatal O3 exposure significantly increased lung
inflammatory cytokine levels; this was further
exacerbated with gestational PM exposure.
Non-significant increases airway resistance and
airway responsiveness with O3 or inhaled allergen
alone. Allergen + O3 produced additive changes in
both measures.
Cellular changes and significant structural
changes in the distal respiratory tract in infant
rhesus monkeys exposed to O3 postnatally.
Laterality changes in offspring: Ozone exposed
pups showed a turning preference (left turns)
distinct from air exposed controls (clockwise turns)
as adults.
Developmental O3 caused increased
defensive/submissive behavior in offspring. Ozone
GD17.
exposed offspring also had significant elevations
of striatal BDNF and hippocampal NGF v. air
exposed controls.
7-73
-------
Study
Han et al.
(2011)
Campos-
Bedolla et
al. (2002)
Kavlock et
al. (1980)
Jedlinska-
Krakowska
et al. (2006)
Model
Rat; Sprague
Dawley, M &
F; PND13
Pregnant
Rats; Sprague
Dawley (GD5,
GD10, or
GD18)
CD-1 mice;
(pregnancy
day 7-1 7)
5 month old
male Wistar
Hannover rats
03
(ppm)
0.6
3.0
0.4,
0.8
and
1.2
3.0
Exposure Duration
3 h, BALF examined
10h after O3
exposure
1 h on one day of
gestation, uteri
collected 16-18 h
later
Continuous,
pregnancy day 7-17
0.5 ppm, 5h/day for
50 days
Effects
BALF polymorphonuclear leukocytes and total
BALF protein were significantly elevated in O3
exposed pups. Lung tissue from O3 exposed pups
had significant elevations of manganese
superoxide dismutase (SOD) protein and
significant decrements of extra-cellular SOD
protein.
Ozone inhalation modifies the contractile response
of the pregnant uterus. The O3 exposed pregnant
uteri had significant increases in the maximum
response to acetyl choline stimulation at GD5 and
10; they also had a significant increase in maximal
response to oxytocin at GD 5.
O3 induced decrements in postnatal body weight
gain. When O3 was co-administered with sodium
salicylate, O3 synergistically increased the rate of
pup resorption (1.0 ppm GD9-12).
Histopathological evidence of impaired
spermatogenesis (round spermatids/21
spermatocytes, giant spermatid cells, and focal
epithelial desquamation with denudation to the 22
basement membrane). Vitamin E exposure
concomitant with O3 protected against
pathological changes but Vitamin C did not.
7.4.11 Summary and Causal Determination
The 2006 O3 AQCD concluded that the limited number of studies that investigated
O3 demonstrated no associations between O3 and birth outcomes, with the possible
exception of birth defects. The current review included an expanded body of
evidence on the associations between O3 and reproductive and developmental
effects. Recent epidemiologic and toxicological studies provide evidence for an
effect of prenatal exposure to O3 on pulmonary structure and function, including lung
function changes in the newborn, incident asthma, ultrastructural changes in
bronchiole development, alterations in placental and pup cytokines, and increased
pup airway hyper-reactivity. Also, there is limited toxicological evidence for an
effect of prenatal and early life exposure on central nervous system effects, including
laterality, brain morphology, neurobehavioral abnormalities, and sleep aberration.
Recent epidemiologic studies have begun to explore the effects of O3 on sperm
quality, and provide limited evidence for decrements in sperm concentration, while
there is limited toxicological evidence for testicular degeneration associated with O3.
While the collective evidence for many of the birth outcomes examined is generally
inconsistent (including birth defects), there are several well-designed, well-conducted
studies that indicate an association between O3 and adverse outcomes. For example,
as part of the southern California Children's Health Study, Salam et al. (2005)
7-74
-------
observed a concentration-response relationship of decreasing birth weight with
increasing O3 concentrations averaged over the entire pregnancy that was clearest
above the 30-ppb level (see Figure 7-4). Similarly, Hansen et al. (2008) utilized fetal
ultrasonic measurements and found a change in ultrasound measurements associated
with O3 during days 31-60 of gestation indicated that increasing O3 concentration
decreased an ultrasound measurement for women living within 2 km of the
monitoring site.
The weight of evidence does not indicate that prenatal or early life O3 concentrations
are associated with infant mortality. Collectively, there is limited though positive
toxicological evidence for O3-induced developmental effects, including effects on
pulmonary structure and function and central nervous system effects. Limited
epidemiologic evidence for an effect on prenatal O3 exposure on respiratory
development provides coherence with the effects observed in toxicological studies.
There is also limited epidemiologic evidence for an association with O3
concentration and decreased sperm concentration. A recent toxicological study
provides limited evidence for a possible biological mechanism (histopathology
showing impaired spermatogenesis) for such an association. Additionally, though the
evidence for an association between O3 concentrations and adverse birth outcomes is
generally inconsistent, there are several influential studies that indicate an association
with reduced birth weight and restricted fetal growth.
Some of the key challenges to interpretation of these study results include the
difficulty in assessing exposure as most studies use existing monitoring networks to
estimate individual exposure to ambient air pollution (see Section 4.6): the inability
to control for potential confounders such as other risk factors that affect birth
outcomes (e.g., smoking); evaluating the exposure window (e.g., trimester) of
importance; integrating the results from both short- and long-term exposure periods;
integrating the results across a variety of reproductive and developmental outcomes;
and limited evidence on the physiological mechanism of these effects.
Taking into consideration the positive evidence for developmental and reproductive
outcomes from toxicological and epidemiological studies, and the few influential
birth outcome studies, the evidence is suggestive of a causal relationship
between exposures to O3 and reproductive and developmental effects.
7.5 Central Nervous System Effects
7.5.1 Effects on the Brain and Behavior
The 2006 O3 AQCD (U.S. EPA. 2006b) included toxicological evidence that acute
exposures to O3 are associated with alterations in neurotransmitters, motor activity,
short and long term memory, and sleep patterns. Additionally, histological signs of
neurodegeneration have been observed. Reports of headache, dizziness, and irritation
7-75
-------
of the nose with O3 exposure are common complaints in humans, and some
behavioral changes in animals may be related to these symptoms rather than
indicative of neurotoxicity. Research in the area of O3-induced neurotoxicity has
notably increased over the past few years, and recent studies examining the effects of
long-term exposure have demonstrated progressive damage in various regions of the
brains of rodents in conjunction with altered behavior. Evidence from epidemiologic
studies has been more limited. A recently published epidemiologic study examined
the association between O3 concentration and neurobehavioral effects. Chen and
Schwartz (2009) utilized data from the NHANES III cohort to study the relationship
between O3 concentrations (mean annual O3 concentration 26.5 ppb) and
neurobehavioral effects among adults aged 20-59 years. Annual O3 concentration
was determined using inverse distance weighting for county of residence and
adjacent counties (for more information on inverse distance weighting and other
methods for exposure assessment, see Sections 4.5.1 and 4.6). The authors observed
an association between annual O3 concentration and tests measuring coding ability
(symbol-digit substitution test) and attention/short-term memory (serial-digit learning
test). Each 10-ppb increase in annual O3 concentration corresponded to an aging-
related cognitive performance decline of 3.5 yr for coding ability and 5.3 years for
attention/short-term memory. These associations persisted in both crude and adjusted
models. There was no association between O3 concentration and reaction time tests.
The authors concluded that overall, there is an association between long-term O3
concentration and reduced performance on neurobehavioral tests.
A number of recent toxicological studies demonstrate various perturbations in
neurologic function or histology with long-term exposure to O3, including changes
similar to those observed in neurodegenerative disorders such as Parkinson's and
Alzheimer's disease pathologies in relevant regions of the brain (Table 7-11).
The central nervous system is very sensitive to oxidative stress, due in part to its high
content of polyunsaturated fatty acids, high rate of oxygen consumption, and low
antioxidant enzyme capacity. Oxidative stress has been identified as one of the
pathophysiological mechanisms underlying neurodegenerative disease (Simonian and
Coyle. 1996). and it is believed to play a role in altering hippocampal function,
which causes cognitive deficits with aging (Vanguilder and Freeman. 2011).
A particularly common finding in studies of O3-exposed rats is lipid peroxidation in
the brain, especially in the hippocampus, which is important for higher cognitive
function including contextual memory acquisition. Performance in passive avoidance
learning tests is impaired when the hippocampus is injured. For example, in a
subchronic study, exposure of rats to 0.25 ppm O3 (4 h/day) for 15-90 days caused a
complex array of responses, including a time-dependent increase in lipid
peroxidation products and immunohistochemical changes in the hippocampus that
were correlated with decrements in passive avoidance behavioral tests (Rivas-
Arancibia et al.. 2010). Changes included increased numbers of activated microglia,
a sign of inflammation, and progressive neurodegeneration. Notably, continued
exposure tends to bring about progressive, cumulative damage, as shown by this
study (Rivas-Arancibia et al.. 2010) and others (Santiago-Lopez et al.. 2010:
Guevara-Guzman et al., 2009; Angoa-Perez et al., 2006). The effects of O3 on
passive avoidance test performance were particularly evident at 90 days for both
7-76
-------
short- and long-term memory. The greatest extent of cell loss was also observed at
this time point, whereas lipid peroxidation did not increase much beyond 60 days of
exposure.
The substantia nigra is another region of the brain affected by O3, and seems
particularly sensitive to oxidative stress because the metabolism of dopamine, central
to its function, is an oxidative process perturbed by redox imbalance. Oxidative stress
has been implicated in the premature death of substantia nigra dopamine neurons in
Parkinson's disease. Progressive damage has been found in the substantia nigra of
male rats after 15, 30, and 60 days of exposure to 0.25 ppm O3 for 4 h/day. Santiago-
Lopez et al. (2010) observed a reduction dopaminergic neurons within the substantia
nigra over time, with a complete loss of normal morphology in the remaining cells
and virtually no dopamine immunoreactivity at 60 days. This was accompanied by an
increase in p53 levels and nuclear translocation, a process associated with
programmed cell death. Similarly, Angoa-Perez et al. (2006) have shown progressive
lipoperoxidation in the substantia nigra and a decrease in nigral neurons in
ovariectomized female rats exposed to 0.25 ppm O3, 4h/day, for 7-60 days. Lipid
peroxidation effectively doubled between the 30 and 60 day time points. Total nigral
cell number was also diminished to the greatest extent at 60 days, and cell loss was
particularly evident in the tyrosine hydroxylase positive cell population (90%),
indicating a selective loss of dopamine neurons or a loss of dopamine pathway
functionality.
The olfactory bulb also undergoes oxidative damage in O3-exposed animals, in some
cases altering olfactory-dependent behavior. Lipid peroxidation was observed in the
olfactory bulbs of ovariectomized female rats exposed to 0.25 ppm O3 (4 h/day) for
30 or 60 days (Guevara-Guzman et al.. 2009). Ozone also induced decrements in a
selective olfactory recognition memory test, which were significantly greater at
60 days compared to 30 days, and the authors note that early deficits in odor
perception and memory are components of human neurodegenerative diseases.
The decrements in olfactory memory did not appear to be due to damaged olfactory
perception based on other tests early on, but by 60 days deficits in olfactory
perception had emerged.
Memory deficits and associated morphological changes can be attenuated by
administration of a-tocopherol (Guerrero et al.. 1999). taurine (Rivas-Arancibia et
al.. 2000). and estradiol (Guevara-Guzman et al.. 2009: Angoa-Perez et al.. 2006). all
of which have antioxidant properties. In the study by Angoa-Perez et al. (2006)
described above, estradiol seemed particularly effective at protecting against lipid
peroxidation and nigral cell loss at 60 days compared to shorter exposure durations.
The same was true for amelioration of decrements in olfactory recognition memory
(Guevara-Guzman et al.. 2009). although protection against lipid peroxidation was
similar for the 30 and 60 day exposures.
CNS effects have also been demonstrated in adult mice whose only exposure to O3
occurred while in utero, a period particularly critical for brain development. Santucci
et al. (2006) investigated behavioral effects and gene expression after in utero
exposure of mice to 0.3 or 0.6 ppm O3. Exposure began 30 days prior to mating and
7-77
-------
continued throughout gestation. Testing of adult animals demonstrated increased
defensive/submissive behavior and reduced social investigation in both the 0.3 and
0.6 ppm O3 groups. Changes in gene expression of brain-derived neurotrophic factor
(BDNF, increased in striatum) and nerve growth factor (NGF, decreased in
hippocampus) accompanied these behavioral changes. BDNF and NGF are involved
in neuronal organization and the growth, maintenance, and survival of neurons
during early development and in adulthood. This study and two others using short-
term exposures demonstrate that CNS effects can occur as a result of in utero
exposure to O3, and although the mode of action of these effects is not known, it has
been suggested that circulating lipid peroxidation products may play a role
(Boussouar et al. 2009). Importantly, these CNS effects occurred in rodent models
after in utero only exposure to (semi-) relevant concentrations of O3.
7-78
-------
Table 7-11 Central nervous system effects of long-term O$ exposure in rats.
Study
Model
O3 (ppm) Exposure Duration Effects
Anqoa-Perez
et al. (2006)
Rat; Wistar; F;
Weight: 300 g;
Ovariectomized
0.25 7 to 60 days,
4 h/day, 5 days/week
Long-term estradiol treatment
protected against Os-induced
oxidative damage to nigral
dopamine neurons, lipid
peroxidation, and loss of tyrosine
hydrolase-immunopositive cells.
Guevara- Rat; Wistar; F; 0.25 30 and 60 days, Long-term estradiol treatment
Guzman et al. Weight: 264 g; 4h/day protected against Os-induced
(2009) Ovariectomized oxidative stress and decreases in a
and (3 estrogen receptors and
dopamine (3-hydroxlyase in olfactory
bulb, and deficits in olfactory social
recognition memory and chocolate
recognition.
Rivas-
Arancibia et al.
Rat; Wistar; M;
Weight: 250-300 g
0.25 15 to 90 days, Ozone produced significant
4h/day increases in lipid peroxidation in the
hippocampus, and altered the
number of p53 positive
immunoreactive cells, activated and
phagocytic microglia, GFAP
immunoreactive cells, double cortine
cells, and short- and long-term
memory-retention latency
Santiago-
Lopez et al.
(2010)
Rat; Wistar; M;
Weight: 250-300 g
0.25 15, 30, and 60 days,
4 h/day
Progressive loss of dopamine
reactivity in the substantia nigra,
along with morphological changes.
Increased p53 levels and nuclear
translocation.
Santucci et al. Mice; CD-1; M; 0.3; 0.6 Females
(2006) 18 weeks old continuously
exposed from 30
days prior to
breeding until GD17
Upon behavioral challenge with
another male, there was a
significant increase in defensive and
freezing postures and decrease in
the frequency of nose-sniffing.
These behavioral changes were
accompanied by a significant
increase in BDNF in the striatum
and a decrease of NGF in the
hippocampus.
7.5.2 Summary and Causal Determination
The 2006 O3 AQCD included toxicological evidence that acute exposures to O3 are
associated with alterations in neurotransmitters, motor activity, short and long term
memory, and sleep patterns. Additionally, histological signs of neurodegeneration
have been observed. However, evidence regarding chronic exposure and
neurobehavioral effects was not available. Recent research in the area of O3-induced
neurotoxicity has included several long-term exposure studies. Notably, the first
7-79
-------
epidemiologic study to examine the relationship between O3 exposure and
neurobehavioral effects observed an association between annual O3 levels and an
aging-related cognitive performance decline in tests measuring coding ability and
attention/short-term memory. This observation is supported by studies in rodents
which demonstrate progressive oxidative stress and damage in the brain and
associated decrements in behavioral tests, including those measuring memory, after
subchronic exposure to 0.25 ppm O3. Additionally, neurobehavioral changes are
evident in animals whose only exposure to O3 occurred in utero. Collectively, the
limited epidemiologic and toxicological evidence is coherent and suggestive of a
causal relationship between O3 exposure and CMS effects.
7.6 Carcinogenic and Genotoxic Potential of Ozone
7.6.1 Introduction
The radiomimetic and clastogenic qualities of O3, combined with its ability to
stimulate proliferation of cells in the respiratory tract, have suggested that O3 could
act as a carcinogen. However, toxicological studies of tumorigenesis in the rodent
lung have yielded mixed and often confusing results, and the epidemiologic evidence
is equally conflicted. The 2006 O3 AQCD concluded that, "the weight of evidence
from recent animal toxicological studies and a very limited number of epidemiologic
studies do not support ambient O3 as a pulmonary carcinogen"1 (U.S. EPA, 2006b).
Multiple epidemiologic studies reported in the 2006 O3 AQCD examined the
association between O3 concentration and cancer. The largest of these studies, by
Pope et al. (2002). included 500,000 adults from the American Cancer Society's
(ACS) Cancer Prevention II study. In this study, no association was observed
between O3 concentration and lung cancer mortality. The Adventist Health Study of
Smog (AHSMOG) also examined the association between O3 concentration and lung
cancer mortality (Abbey et al.. 1999). There was a positive association between O3
concentrations and lung cancer mortality among men. No association was reported
for women. Another study using the AHSMOG cohort assessed the risk of incident
lung cancer (Beeson et al., 1998). Among males, an association with incidence of
lung cancer was observed with increasing O3 concentrations. When stratified by
smoking status, the association persisted among never smokers but was null for
former smokers. No association was detected for females. The Six Cities Study
examined various air pollutants and mortality but did not specifically explore the
association between O3 concentrations and lung cancer mortality due to low
variability in O3 concentrations across the cities (Dockery et al., 1993). An ecologic
study performed in Sao Paulo City, Brazil examined the correlations between O3
concentrations in four of the city districts and incident cancer of the larynx and lung
1 The toxicological evidence is presented in detail in Table 6-18 on page 6-116 of the 1996 O3AQCD (U.S. EPA. 1996a)and
Table AX5-1 Son page AX5-43 of the 2006 O3 AQCD (U.S. EPA. 2006b).
7-80
-------
reported in 1997 (Pereira et al., 2005). A correlation between the average number of
days O3 concentrations exceeded air quality standards from 1981 to 1990 and cancer
incidence was present for larynx cancer but not for lung cancer.
Early toxicological research demonstrated lung adenoma1 acceleration in mice with
daily exposure to 1 ppm over 15 months (Stokinger. 1962). Later work demonstrated
a significant increase in lung tumor numbers in one strain of mouse (A/J) but not
another after exposure to 0.3-0.8 ppm O3 (Lastetal, 1987; Hassett et al., 1985).
The A/J mouse strain is known to have a high incidence of spontaneous adenomas,
and further studies using this strain found a statistically significant increase in lung
tumor incidence after a 9-month exposure to 0.5 ppm and incidence and multiplicity
after a 5 month exposure to 0.12 ppm with a 4-month recovery period (Witschi et al.,
1999). However, these findings were discounted by the study authors due to the lack
of a clear concentration-response, and results from the Hassett et al. (1985) and Last
et al. (1987) studies were retrospectively deemed spurious based on what appeared to
be unusually low spontaneous tumor incidences in the control groups (Witschi,
1991). A study of carcinogenicity of O3 by the National Toxicology Program (NTP,
1994) reported increased incidences of alveolar/bronchiolar adenoma or carcinoma
(combined) in female B6C3Fi mice exposed over 2 years to 1.0 ppm O3, but not
0.12 or .5 ppm. No effect was detected in male mice. For a lifetime exposure to 0.5
or 1.0 ppm O3, an increase in the number of female mice with adenomas (but not
carcinomas or total neoplasms) was found. The number of total neoplasms was also
unaffected in male mice, but there was a marginally increased incidence of
carcinoma in males exposed to 0.5 and 1.0 ppm. Thus there was equivocal evidence
of carcinogenic activity in male mice and some evidence of carcinogenic activity of
O3 in females. Experimental details of the NTP mouse study are available in
Table 6-19 on page 6-121 (U.S. EPA. 1996o) of the 1996 O3 AQCD (U.S. EPA.
1996a).
In Fischer-344/N rats (50 of each sex per group), neither a 2-year nor lifetime
exposure to O3 ranging from 0.12 to 1.0 ppm was found to be carcinogenic
(Boorman et al., 1994; NTP, 1994). However, a marginally significant carcinogenic
effect of 0.2 ppm O3 was reported in a study of male Sprague-Dawley rats exposed
for 6 months (n = 50) (Monchaux et al., 1996). These two studies also examined co-
carcinogenicity of O3 with NNK2 (Boorman et al., 1994) or a relatively high dose of
radon (Monchaux et al., 1996), finding no enhancement of NNK related tumors and a
slight non-significant increase in tumor incidence after combined exposure with
radon, respectively. Another study exploring co-carcinogenicity was conducted in
hamsters. Not only was there no enhancement of chemically induced tumors in the
peripheral lung or nasal cavity, but results suggested that O3 could potentially delay
or inhibit tumor development (Witschi et al., 1993). Thus there is no concrete
evidence that O3 can act as a co-carcinogen.
1 NOTE: Although adenomas are benign, over time they may progress to become malignant, at which point they are called
adenocarcinomas. Adenocarcinoma is the predominant lung cancer subtype in most countries, and is the only lung cancer found
in nonsmokers. From page 8-33 of the 1970 OjAQCD: "No true lung cancers have been reported, however, from experimental
exposures to either O3 alone or any other combination or ingredient of photochemical oxidants."
2 4-(N-nitrosomethylamino)-1 -(3-pyridyl)-1 -butanone
7-81
-------
Immune surveillance is an important defense against cancer, and it should be noted
that natural killer (NK) cells, which destroy tumor cells in the lung, appear to be
inhibited by higher concentrations of O3 and either unaffected or stimulated at lower
concentrations (Section 6.2.5.4, Infection and Adaptive Immunity). This aspect of
tumorigenesis adds yet another layer of complexity which may be reflected by
conflicting results across studies.
The following sections will examine epidemiologic studies of cancer incidence and
mortality and toxicological studies that have been published since the 2006 O3
AQCD. An epidemiologic study has been published with cancer as the outcome;
most epidemiologic studies examine markers of exposure.
7.6.2 Lung Cancer Incidence and Mortality
A recent re-analysis of the full ACS CPSII cohort by the Health Effects Institute is
the only epidemiologic study that has explored the association between O3
concentration and cancer mortality since the last O3 AQCD. Krewski et al. (2009)
conducted an extended follow-up of the cohort (1982-2000). Mean O3 concentration
[obtained from the Aerometric Information Retrieval System (AIRS) for 1980] were
22.91 ppb for the full year and 30.15 ppb for the summer months (April-September).
No association was reported between lung cancer mortality and O3 concentration
(HR = 1.00 [95% CI: 0.96-1.04] per 10 ppb O3). Additionally, no association was
observed when the analysis was restricted to the summer months. There was also no
association present in a sub-analysis of the cohort examining the relationship
between O3 concentration and lung cancer mortality in the Los Angeles area.
Since the 2006 O3 AQCD, two toxicological studies have examined potential
carcinogenicity of O3 (Kim and Cho. 2009a. b). Looking across both studies, which
used the same mouse strain as the National Toxicology Program study described
above (NTP. 1994). 0.5 ppm O3 alone or in conjunction with chemical tumor
inducers did not enhance lung tumor incidence in males or females. However, a 10%
incidence of oviductal carcinoma was observed in mice exposed to 0.5 ppm O3 for
16 weeks. The implications of this observation are unclear, particularly in light of the
lack of statistical information reported. Additionally, there is no mention of oviductal
carcinoma after 32 weeks of exposure, and no oviductal carcinoma was observed
after one year of exposure. The NTP study did not report any increase in tumors at
extrapulmonary sites.
7.6.3 DMA Damage
The potential for genotoxic effects relating to O3 exposure was predicted from the
radiomimetic properties of O3. The decomposition of O3 in water produces OH and
HO2 radicals, the same species that are generally considered to be the biologically
active products of ionizing radiation. Ozone has been observed to cause degradation
7-82
-------
of DNA in a number of different models and bacterial strains. The toxic effects of O3
have been generally assumed to be confined to the tissues directly in contact with the
gas, such as the respiratory epithelium. Due to the highly reactive nature of O3, little
systemic absorption is predicted. Zelac et al. (1971 a, b); however, reported a
significant increase in chromosome aberrations in peripheral blood lymphocytes
from Chinese hamsters exposed to 0.2 ppm for 5 hours. Other in vivo exposure
studies found increased DNA strand breaks in respiratory cells from guinea pigs
(Ferng et al.. 1997) and mice (Bornholdt et al.. 2002) but only with exposure to
higher concentrations of O3 (1 ppm for 72 hours and 1 or 2 ppm for 90 minutes,
respectively). In other studies there were no observations of chromosomal
aberrations in germ cells, but mutagenic effects have been seen in offspring of mice
exposed to 0.2 ppm during gestation (blepharophimosis or dysplasia of the eyelids).
The overall evidence for mutagenic activity from in vitro studies is positive, and in
the National Toxicology Program report described above, O3 was found to be
mutagenic in Salmonella, with and without S9 metabolic activation. No recent
toxicological studies of DNA damage have become available since the 2006 O3
AQCD.
A number of epidemiologic studies looked at the association between O3 and DNA
and cellular level damages. These changes may be relevant to mechanisms leading to
cancers development and serve as early indicators of elevated risk of mutagenicity.
Two studies performed in California examined cytogenetic damage in relation to O3
exposures. Huen et al. (2006) examined cytogenetic damage among African
American children and their mothers in Oakland, CA. Increased O3 (mean monthly
8-h O3 concentrations ranged from about 30 ppb in April to 14 ppb in November)
was associated with increased cytogenetic damage (micronuclei frequency among
lymphocytes and buccal cells) even after adjustment for household/personal smoking
status and distance-weighted traffic density. Chen et al. (2006a) recruited college
students at the University or California, Berkeley who reported never smoking and
compared their levels of cytogenetic damage (micronuclei frequency from buccal
cells) in the spring and fall. Cytogenetic damage was greater in the fall, which the
authors attributed to the increase in O3 over the summer. However, O3 levels over 2,
7, 10, 14, or 30 days (concentrations not given) before collection of buccal cells did
not correlate with cytogenetic damage. Estimated lifetime O3 exposure was also not
correlated with cytogenetic damage. Additionally, the authors exposed a subset of the
students (n = 15) to 200 ppb O3 for 4 hours while the students exercised
intermittently. Ozone was found to be associated with an increase in cytogenetic
damage in degenerated cells but not in normal cells 9-10 days after exposure.
Increased cytogenetic damage was also noted in peripheral blood lymphocytes
collected 18 hours after exposure.
A study performed in Mexico recruited 55 male workers working indoors (n = 27) or
outdoors (n = 28) in Mexico City or Puebla, Mexico in order to study the relationship
between O3 and DNA damage (detected from peripheral blood samples using the
Comet assay) (Tovalin et al.. 2006). The median estimated daily O3 concentrations
were estimated to be 28.5 ppb for outdoor workers and 5.1 ppb for indoor workers in
7-83
-------
Mexico City and 36.1 ppb for outdoor workers and 19.5 ppb for indoor workers in
Puebla. Overall, a positive correlation between O3 levels and DNA damage was
observed. However, when examining the relationship by city and workplace, only
DNA damage in outdoor workers in Mexico City remained correlated with O3 levels.
Three studies examining the relationship between O3 concentration and DNA-level
damage have been performed in Europe. The largest of these studies was the GenAir
case-control study, which was nested within the European Prospective Investigation
into Cancer and Nutrition (EPIC) study, and included individuals recruited between
1993 and 1998 from ten European countries. Only non-smokers (must not have
smoked for at least 10 years prior to enrollment) were enrolled in the study.
The researchers examined DNA adduct levels (DNA bonded to cancer-causing
chemicals) and their relationship with O3 concentrations (concentrations not given)
(Peluso et al., 2005). A positive association was seen between DNA adduct levels
and O3 concentrations from 1990-1994 but not O3 concentrations from 1995-1999.
In adjusted analyses with DNA adduct levels dichotomized as high and low
(detectable versus non-detectable), the OR was 1.97 (95% CI: 1.08, 3.58) when
comparing the upper tertile of O3 concentration to the lower two tertiles. Two other
European studies were conducted in Florence, Italy. The most recent of these
enrolled individuals from the EPIC study into a separate study between March and
September of 1999 (Palli et al., 2009). The purpose of the study was to examine
oxidative DNA damage (determined by Comet assay using blood lymphocytes) in
association with varying periods of O3 exposure. The researchers observed that
longer periods of high O3 concentrations (values not given) were more strongly
correlated with oxidative DNA damage than shorter periods of time (i.e., the rho [p-
value] was 0.26 [0.03] for 0-10 days and 0.35 [0.002] for 0-90 days). This correlation
was stronger among men compared to women. The correlations for all time periods
had p-values <0.05 for ex- and never-smokers. For current smokers, the correlation
was only observed among time periods < 25 days. When adjusted for age, sex,
smoking history, traffic pollution exposure, period of blood draw, and area of
residence, the association between O3 concentrations and oxidative DNA damage
was positive for O3 concentrations 0-60 days, 0-75 days, and 0-90 days prior to
blood draw. Positive, statistically significant associations were not observed among
shorter time periods. The other study performed in Florence recruited healthy
volunteers who reported being non-smokers or light smokers (Giovannelli et al..
2006). The estimated O3 concentrations during the study ranged from approximately
4-40 ppb for 3-day averages, 5-35 ppb for 7-day averages, and 7.5-32.5 ppb for 30-
day averages. Ozone concentrations were correlated with DNA strand breaks
(measured from blood lymphocytes) over longer exposure periods (p-value: 0.002 at
30 days, p-value: 0.04 at 7 days; p-value: 0.17 at 3 days). This association was robust
to control for temperature, solar radiation, sex, and age. No association was seen
between O3 concentrations and measures of oxidative DNA damage at 3, 7, or
30 days.
7-84
-------
7.6.4 Summary and Causal Determination
The 2006 O3 AQCD reported that evidence did not support ambient O3 as a
pulmonary carcinogen. Since the 2006 O3 AQCD, very few epidemiologic and
toxicological studies have been published that examine O3 as a carcinogen, but
collectively, study results indicate that O3 may contribute to DNA damage. Ozone
concentrations in most epidemiologic studies were measured using air monitoring
data. For more information on long-term exposure assessment, see Section 4.6.3.2.
Overall, the evidence is inadequate to determine if a causal relationship exists
between ambient O3 exposures and cancer.
7.7 Mortality
A limited number of epidemiologic studies have assessed the relationship between
long-term exposure to O3 and mortality in adults. The 2006 O3 AQCD concluded
that an insufficient amount of evidence existed "to suggest a causal relationship
between chronic O3 exposure and increased risk for mortality in humans" (U.S. EPA.
2006b). In addition to the infant mortality studies discussed in Section 7.4.10.
additional studies have been conducted among adults since the last review; an
ecologic study that finds no association between mortality and O3, several re-
analyses of the ACS cohort, one of which specifically points to a relationship
between long-term O3 exposure and an increased risk of respiratory mortality, and a
study of four cohorts of persons with potentially predisposing conditions. These
studies supplement the evidence from long-term cohort studies characterized in
previous reviews of O3, and are summarized here briefly.
In the Harvard Six Cities Study (Dockery et al.. 1993). adjusted mortality rate ratios
were examined in relation to long-term mean O3 concentrations in six cities: Topeka,
KS; St. Louis, MO; Portage, WI; Harriman, TN; Steubenville, OH; and Watertown,
MA. Mean O3 concentrations from 1977 to 1985 ranged from 19.7 ppb in Watertown
to 28.0 ppb in Portage. Long-term mean O3 concentrations were not found to be
associated with mortality in the six cities. However, the authors noted that
"The small differences in O3 levels among the (six) cities limited the power of the
study to detect associations between mortality and O3 levels." In addition, while total
and cardio-pulmonary mortality were considered in this study, respiratory mortality
was not specifically considered.
In a subsequent large prospective cohort study of approximately 500,000 U.S. adults,
Pope et al. (2002) examined the effects of long-term exposure to air pollutants on
mortality (American Cancer Society, Cancer Prevention Study II). All-cause,
cardiopulmonary, lung cancer and other mortality risk estimates for long-term O3
exposure are shown in Figure 7-5. While consistently positive associations were not
observed between O3 and mortality (effect estimates labeled "A" in Figure 7-5). the
mortality risk estimates were larger in magnitude when analyses considered more
accurate exposure metrics, increasing when the entire period was considered (effect
7-85
-------
estimates labeled "B" in Figure 7-5) and becoming marginally significant when the
exposure estimate was restricted to the summer months (July to September; effect
estimates labeled "C" in Figure 7-5), especially when considering cardiopulmonary
deaths. In contrast, consistent positive and significant effects of PM2.5 were observed
for both lung cancer and cardio-pulmonary mortality.
All Cause Cardiopulmonary
Mortality Mortality Lung Cancer All Other Causes
_ Mortality Mortalilv
u
cc
cr
0
o
1
A B C A B C A 13 C A B C
Number of Number of Participants
Years of Data Collection Metropolitan Areas (in thousands) 1-h max O3 Mean (SD)
A 1980-1981 134 559 47.9(11.0)
B 1982-1998 119 525 45.5(7.3)
C 1982-1998 (July-Sept) 134 557 59.7(12.8)
Source: Reprinted with permission of American Medical Association Pope et al. (2002).
Figure 7-5 Adjusted O3-mortality relative risk estimates (95% Cl) by time
period of analysis per subject-weighted mean O3 concentration in
the Cancer Prevention Study II by the American Cancer Society.
A study by Abbey et al. (1999) examined the effects of long-term air pollution
exposure, including O3, on all-cause (n = 1,575), cardiopulmonary (n = 1,029),
nonmalignant respiratory (n = 410), and lung cancer (n = 30) mortality in the long-
term prospective Adventist Health Study of Smog (AHSMOG) of 6,338 nonsmoking,
non-Hispanic white individuals living in California. A particular strength of this
study was the extensive effort devoted to assessing long-term air pollution exposures,
including interpolation to residential and work locations from monitoring sites over
time and space. No associations with long-term O3 exposure were observed for all
cause, cardiopulmonary, and nonmalignant respiratory mortality. In a follow-up,
Chen et al. (2005) utilized data from the AHSMOG study and reported no evidence
of associations between long-term O3 exposure (mean O3 concentration 26.2 ppb)
7-86
-------
and fatal coronary heart disease. Thus, no association of chronic O3 exposure with
mortality outcomes has been detected in this study.
Lipfert et al. (2003. 2000) reported positive effects on all-cause mortality for peak O3
exposures (95th percentile levels) in the U.S. Veterans Cohort study of
approximately 50,000 middle-aged men recruited with a diagnosis of hypertension.
The actual analysis involved smaller subcohorts based on exposure and mortality
follow-up periods. Four separate exposure periods were associated with three
mortality follow-up periods. For concurrent exposure periods, peak O3 was
positively associated with all-cause mortality, with a 9.4% (95% CI: 0.4, 18.4) excess
risk per mean 95th percentile O3 less estimated background level (not stated). "Peak"
refers, in this case, to the 95th percentile of the hourly measurements, averaged by
year and county. In a further analysis, Lipfert et al. (2003) reported the strongest
positive association for concurrent exposure to peak O3 for the subset of subjects
with low diastolic blood pressure during the 1982 to 1988 period. Two more recent
studies of this cohort focused specifically on traffic density (Lipfert et al., 2006a;
2006b). Lipfert et al. (2006b) concluded that: "Traffic density is seen to be a
significant and robust predictor of survival in this cohort, more so than ambient air
quality, with the possible exception of O3," reporting a significant O3 effect even
with traffic density included in the model: RR = 1.080 per 40 ppb peak O3 (95% CI:
1.019, 1.146). However, in Lipfert et al. (2006a), which considers only the EPA
Speciation Trends Network (STN) sites, O3 drops to non-significant predictor of total
mortality for this cohort. The authors acknowledge that: "Peak O3 has been important
in analyses of this cohort for previous periods, but in the STN data set, this variable
has limited range and somewhat lower values and its small coefficient of variation
results in a relatively large standard error." The restriction to subjects near STN sites
likely reduced the power of this analysis, though the size of the remaining subjects
considered was not reported in this paper. In addition, these various Veterans Cohort
studies considered only total mortality, and did not consider mortality on a by-cause
basis.
An ecological study in Brisbane, Australia used a geospatial approach to analyze the
association of long-term exposure to gaseous air pollution with cardio-respiratory
mortality, in the period 1996-2004 (Wang et al., 2009c). A generalized estimating
equations model was employed to investigate the impact of NO2, O3 and SO2, but
PM was not addressed. The results indicated that long-term exposure to O3 was not
associated with cardio-respiratory mortality, but the fact that this study considered
only one city, and that the range of O3 exposure across that city (23.7-35.6 ppb) was
low and slight in variation in comparison to the range of other pollutants across the
city, limited study power. In addition, confounding factors (e.g., smoking) could not
be addressed at the individual level in this ecological study. Respiratory mortality
was not evaluated separately.
A recent study by Zanobetti and Schwartz examined whether year-to-year variations
in 8-h mean daily O3 concentrations for the summer (May-September) around their
city-specific long-term trend were associated with year-to-year variations in mortality
around its long-term trend. This association was examined among Medicare
7-87
-------
participants with potentially predisposing conditions, including COPD, diabetes,
CHF, and MI, defined as patients discharged alive after an emergency admission for
one of these four conditions. The analyses was repeated in 105 cities using available
data from 1985 through 2006, and the results were combined using methods
previously employed by these authors (Zanobetti et al, 2008; Zanobetti and
Schwartz. 2007). This study design eliminated potential confounding by factors that
vary across city, which is a common concern in most air pollution cohort studies, and
also avoided both confounding by cross-sectional factors that vary by city and the
short-term factors that confound daily time-series studies, but are not present in
annual analyses. The average 8-h mean daily summer O3 concentrations ranged from
15.6 ppb (Honolulu, HI) to 71.4 ppb (Bakersfield, CA) for the 105 cities. The authors
observed associations between yearly fluctuations in summer O3 concentrations and
mortality in each of the four cohorts; the hazard ratios (per 10 ppb increment) were
1.12 (95% CI: 1.06, 1.17) for the CHF cohort, 1.19 (95% CI 1.12, 1.25) for the MI
cohort, 1.14 (95% CI: 1.10, 1.21) for the diabetes cohort, and 1.14 (95% CI: 1.08,
1.19) for the COPD cohort. A key advantage to this study is that fluctuations from
summer to summer in O3 concentrations around long-term level and trend in a
specific city are unlikely to be correlated with most other predicators of mortality
risk; except for temperature, which was controlled for in the regression. Key
limitations of the study were the inability to control for PM2.5, since it was not
reliably measured in these cities until 1999, and the inability to separate specific
causes of death (e.g., respiratory, cardiovascular), since Medicare does not provide
the underlying cause of death.
In the most recent follow-up analyses of the ACS cohort (Jerrett et al., 2009; Smith et
al., 2009a), the effects of long-term exposure to O3 were evaluated alone, as well as
in copollutant models with PM2.s and components of PM2.s. Jerrett et al. (2009)
utilized the ACS cohort with data from 1977 through 2000 (mean O3 concentration
ranged from 33.3 to 104.0 ppb) and subdivided cardiopulmonary deaths into
respiratory and cardiovascular, separately, as opposed to combined into one category,
as was done by Pope et al. (2002). Increases in exposure to O3 were associated with
an elevated risk of death from cardiopulmonary, cardiovascular, ischemic heart
disease, and respiratory causes. Consistent with study hypotheses, inclusion of PM2.s
concentrations measured in 1999-2000 (the earliest years for which it was available)
as a copollutant attenuated the association with O3 for all end points except death
from respiratory causes, for which a significant association persisted (Table 7-12).
The association between increased O3 concentrations and increased risk of death
from respiratory causes was insensitive to the use of a random-effects survival model
allowing for spatial clustering within the metropolitan area and state of residence,
and adjustment for several ecologic variables considered individually. Subgroup
analyses showed that temperature and region of country, but not sex, age at
enrollment, body-mass index, education, or PM2 5 concentration, modified the effects
of O3 on the risk of death from respiratory causes (i.e., risks were higher at higher
temperature, and in the Southeast, Southwest, and Upper Midwest). Ozone threshold
analyses indicated that the threshold model was not a better fit to the data (p >0.05)
than a linear representation of the overall O3-mortality association. Overall, this new
analysis indicates that long-term exposure to PM2 5 increases risk of cardiac death,
-------
while long-term exposure to O3 is specifically associated with an increased risk of
respiratory death, and suggests that combining cardiovascular and respiratory causes
of mortality into one category for analysis may obscure any effect that O3 may have
on respiratory-related causes of mortality.
Table 7-12 Relative risk (and 95% Cl) of death attributable to a 10-ppb change
in the ambient O3 concentration.
Cause of Death O3 (96 MSAs)a O3 (86 MSAsf O3 +PM2.5 (86 MSAs)a
Any Cause 1.001 (0.996, 1.007) 1.001 (0.996, 1.007) 0.989 (0.981,0.996)
Cardiopulmonary 1.014(1.007, 1.022) 1.016(1.008, 1.024) 0.992 (0.982, 1.003)
Respiratory 1.029(1.010,1.048) 1.027(1.007,1.046) 1.040(1.013,1.067)
Cardiovascular 1.011(1.003, 1.023) 1.014(1.005, 1.023) 0.983 (0.971,0.994)
Ischemic Heart Disease 1.015(1.003, 1.026) 1.017(1.006, 1.029) 0.973 (0.958, 0.988)
aOzone concentrations were measured from April to September during the years from 1977 to 2000, with follow-up from 1982 to
2000; changes in the concentration of PM25 of 10 ug/m3 were recorded for members of the cohort in 1999 and 2000.
Source: Reprinted with permission of Massachusetts Medical Society (Jerrett et al.. 2009).
In a similar analysis, Smith et al. (2009a) used data from 66 Metropolitan Statistical
Areas (MSAs) in the ACS cohort to examine the association of O3 concentrations
during the warm season and all-cause and cardiopulmonary mortality. Mortality
effects were estimated in single pollutant and copollutant models, adjusting for two
PM2.5 constituents, sulfate, and EC. When all-cause mortality was investigated, there
was a 0.8% (95% CI: -0.31, 1.9) increase associated with a 10 ppb increase in O3
concentration. This association was diminished when sulfate or EC were included in
the model. There was a 2.48% (95% CI: 0.74, 4.3) increase in cardiopulmonary
mortality associated with a 10 ppb increase in O3 concentration.
The cardiopulmonary association was robust to adjustment for sulfate, and
diminished, though still positive, after adjustment for EC (1.63% increase; 95% CI:
-0.41, 3.7). Smith et al. (2009a) did not specifically separate out cardiovascular and
respiratory causes of death from the cardiopulmonary category, as was done by
Jerrett et al. (2009).
7.7.1 Summary and Causal Determination
The The 2006 O3 AQCD concluded that an insufficient amount of evidence existed
"to suggest a causal relationship between chronic O3 exposure and increased risk for
mortality in humans" (U.S. EPA, 2006b). Several additional studies have been
conducted since the last review that evaluate cause-specific and total mortality.
An ecologic study conducted in Australia observed no association between
cardiopulmonary mortality and O3 (Wang et al., 2009c). Two reanalyses of the ACS
cohort were conducted; one provides weak evidence for an association with
cardiopulmonary mortality (Smith et al., 2009a) while the other specifically points to
7-89
-------
a relationship between long-term O3 exposure and an increased risk of respiratory
mortality (Jerrett et al., 2009). Most recently, a study of four cohorts of Medicare
enrollees with potentially predisposing conditions observed associations between O3
and total mortality among each of the cohorts (Zanobetti and Schwartz, 2011).
When considering the entire body of evidence, there is limited support for an
association with long-term exposure to ambient O3 and total mortality. There is
inconsistent evidence for an association between long-term exposure to ambient O3
and cardiopulmonary mortality, with several analyses from the ACS cohort reporting
some positive associations (Smith et al., 2009a; Pope et al., 2002) while other studies
reported no association (Wang et al., 2009c; Abbey et al., 1999; Dockery et al.,
1993). The strongest evidence for an association between long-term exposure to
ambient O3 concentrations and mortality is derived from associations reported in the
Jerrett et al. (2009) study for respiratory mortality that remained robust after
adjusting for PM2.s concentrations. Finally, a recent analysis reported associations of
ambient O3 concentrations and total mortality in potentially at-risk populations in the
Medicare Cohort (Zanobetti and Schwartz, 2011), while earlier studies generally
report no associations with total mortality (Lipfert et al., 2006a; Lipfert et al., 2003;
Pope et al.. 2002; Abbey et al.. 1999; Dockery et al.. 1993). Studies of
cardiopulmonary and total mortality provide limited evidence for an association with
long-term exposure to ambient O3 concentrations. The study by Jerrett et al. (2009)
observes an association between long-term exposure to ambient O3 concentrations
and respiratory mortality that remained robust after adjusting for PM2.s
concentrations. Coherence and biological plausibility for this observation is provided
by evidence from epidemiologic, controlled human exposure, and animal
toxicological studies for the effects of short- and long-term exposure to O3 on
respiratory effects (see Sections 6.2 and 7.2). Respiratory mortality is a relatively
small portion of total mortality [about 7.6% of all deaths in 2010 were due to
respiratory causes (Murphy et al.. 2012)1. thus it is not surprising that the respiratory
mortality signal may be difficult to detect in studies of cardiopulmonary or total
mortality. Based on the recent evidence for respiratory mortality along with limited
evidence for total and cardiopulmonary mortality, the evidence is suggestive of a
causal relationship between long-term O3 exposures and total mortality.
7.8 Overall Summary
The evidence reviewed in this chapter describes the recent findings regarding the
health effects of long-term exposure to ambient O3 concentrations. Table 7-13
provides an overview of the causal determinations for each of the health categories
evaluated.
7-90
-------
Table 7-13 Summary of causal determinations for long-term exposures to O3.
Health Category Causal Determination
Respiratory Effects Likely to be a causal relationship
Cardiovascular Effects Suggestive of a causal relationship
Reproductive and Developmental Effects Suggestive of a causal relationship
Central Nervous System Effects Suggestive of a causal relationship
Carcinogenicity and Genotoxicity Inadequate to infer a causal relationship
Total Mortality Suggestive of a causal relationship
7-91
-------
References
Abbey. DE; Nishino. N; McDonnell. WF; Burchette. RJ; Knutsen. SF; Beeson. WL; Yang. JX. (1999). Long-
term inhalable particles and other air pollutants related to mortality in nonsmokers. Am J Respir Crit Care
Med 159: 373-382.
Akinbami. LJ; Lynch. CD; Parker. JD; Woodruff. TJ. (2010). The association between childhood asthma
prevalence and monitored air pollutants in metropolitan areas, United States, 2001-2004. Environ Res 110:
294-301. http://dx.doi.0rg/10.1016/i.envres.2010.01.001
Andersen. ZJ; Raaschou-Nielsen. O; Ketzel. M; Jensen. SS: Hvidberg. M; Loft. S: Tj0nneland. A; Overvad. K;
S0rensen. M. (2012). Diabetes Incidence and Long-Term Exposure to Air Pollution: A cohort study.
Diabetes Care 35: 92-98. http://dx.doi.org/10.2337/dcll-1155
Angoa-Perez. M; Jiang. H; Rodriguez. AI; Lemini. C: Levine. RA; Rivas-Arancibia. S. (2006). Estrogen
counteracts ozone-induced oxidative stress and nigral neuronal death. Neuroreport 17: 629-633.
Arito. H; Uchiyama. I; Yokoyama. E. (1992). Acute effects of ozone on EEG activity, sleep-wakefulness and
heart rate in rats. Ind Health 30: 23-34.
Auten. RL; Gilmour. MI; Krantz. QT; Potts. EN; Mason. SN; Foster. WM. (2012). Maternal diesel inhalation
increases airway hyperreactivity in ozone-exposed offspring. Am J Respir Cell Mol Biol 46: 454-460.
http://dx.doi.org/10.1165/rcmb.2011-0256OC
Auten. RL; Potts. EN; Mason. SN; Fischer. B; Huang. Y; Foster. WM. (2009). Maternal exposure to particulate
matter increases postnatal ozone-induced airway hyperreactivity in juvenile mice. Am J Respir Crit Care
Med 180: 1218-1226. http://dx.doi.org/10.1164/rccm.200901-0116OC
Barry. BE; Miller. FJ; Crapo. JD. (1985). Effects of inhalation of 0.12 and 0.25 parts per million ozone on the
proximal alveolar region of juvenile and adult rats. Lab Invest 53: 692-704.
Barry. BE; Miller. FJ: Crapo. JD. (1983). Alveolar epithelial injury caused by inhalation of 0.25 ppm of ozone.
In SD Lee; MG Mustafa; MA Mehlman (Eds.), The biomedical effects of ozone and related photochemical
oxidants (pp. 299-308). Princeton, NJ: Princeton Scientific Publishers.
Beeson. WL; Abbey. DE; Knutsen. SF. (1998). Long-term concentrations of ambient air pollutants and incident
lung cancer in California adults: Results from the AHSMOG study. Environ Health Perspect 106: 813-823.
Berhane. K; Gauderman. WJ; Stram. DO; Thomas. DC. (2004). Statistical issues in studies of the long-term
effects of air pollution: The Southern California Childrens Health Study. Stat Sci 19: 414-449.
http://dx.doi.org/10.1214/088342304000000413
Bignami. G; Musi. B; DeH'Omo. G; Laviola. G; Alleva. E. (1994). Limited effects of ozone exposure during
pregnancy on physical and neurobehavioral development of CD-I mice. Toxicol Appl Pharmacol 129: 264-
271. http://dx.doi.org/10.1006/taap. 1994.1251
Bobak. M. (2000). Outdoor air pollution, low birth weight, and prematurity. Environ Health Perspect 108: 173-
176.
Boorman. GA; Hailev. R; Grumbein. S; Chou. BJ; Herbert. RA; Goehl. T; Mellick. PW; Rovcroft. JH;
Haseman. JK; Sills. R. (1994). Toxicology and carcinogenesis studies of ozone and ozone 4-(N-
nitrosomethylamino)-l-(3-pyridyl)-l-butanone in Fischer-344/N rats. Toxicol Pathol 22: 545-554.
Bornholdt. J; Dybdahl. M; Vogel U; Hansen. M; Loft. S; Wallin. H. (2002). Inhalation of ozone induces DNA
strand breaks and inflammation in mice. DNA Repair 520: 63-71.
Boussouar. A; Araneda. S; Hamelin. C; Soulage. C; Kitahama. K; Dalmaz. Y. (2009). Prenatal ozone exposure
abolishes stress activation of Fos and tyrosine hydroxylase in the nucleus tractus solitarius of adult rat.
Neurosci Lett 452: 75-78.
7-92
-------
Brauer. M; Lencar. C: Tamburic. L; Koehoorn. M; Demers. P; Karr. C. (2008). A cohort study of traffic-related
air pollution impacts on birth outcomes. Environ Health Perspect 116: 680-686.
http://dx.doi.org/10.1289/ehp.10952
Breton. CV: Salam. MT; Vora. H; Gauderman. WJ: Gilliland. FD. (2011). Genetic variation in the glutathione
synthesis pathway, air pollution, and children's lung function growth. Am J Respir Crit Care Med 183: 243-
248. http://dx.doi.org/10.1164/rccm.201006-0849OC
Calderon-Garciduenas. L; Mora-Tiscareno. A; Fordham. LA; Chung. CJ: Valencia-Salazar. G: Flores-Gomez.
S: Solt. AC: Gomez-del Campo. A; Jardon-Torres. R; Henriquez-Roldan. C: Hazucha. MJ; Reed. W.
(2006). Lung radiology and pulmonary function of children chronically exposed to air pollution. Environ
Health Perspect 114: 1432-1437.
Campos-Bedolla. P; Vargas. MH; Montano. LM. (2002). Effect of acute ozone exposure on pregnant rat uterus
contractile responses. Reprod Toxicol 16: 269-273.
Carbaial-Arroyo. L; Miranda-Soberanis. V: Medina-Ramon. M; Rojas-Bracho. L; Tzintzun. G: Solis-
Gutierrez. P; Mendez-Ramirez. I; Hurtado-Diaz. M; Schwartz. J; Romieu. I. (2011). Effect of PM10 and
O3 on infant mortality among residents in the Mexico City Metropolitan Area: A case-crossover analysis,
19972005. J Epidemiol Community Health 65: 715-721. http://dx.doi.org/10.1136/iech.2009.101212
Carey. SA; Ballinger. CA; Plopper. CG: McDonald. RJ: Bartolucci. AA; Postlethwait. EM; Harkema. JR.
(2011). Persistent rhinitis and epithelial remodeling induced by cyclic ozone exposure in the nasal airways
of infant monkeys. Am J Physiol Lung Cell Mol Physiol 300: L242-L254.
http://dx.doi.org/10.1152/ajplung.00177.2010
Carey. SA; Minard. KR; Trease. LL; Wagner. JG: Garcia. GJ: Ballinger. CA; Kimbell. JS: Plopper. CG: Corlev.
RA; Postlethwait. EM; Harkema. JR. (2007). Three-dimensional mapping of ozone-induced injury in the
nasal airways of monkeys using magnetic resonance imaging and morphometric techniques. Toxicol Pathol
35: 27-40. http://dx.doi.org/10.1080/01926230601072343
Carter. EP; Garat. C: Imamura. M. (2004). Continual emerging roles of HO-1: protection against airway
inflammation [Review]. Am J Physiol Lung Cell Mol Physiol 287: L24-L25.
http://dx.doi.org/10.1152/ajplung.00097.2004
Catalano. PJ; Chang. LY; Harkema. JR; Kaden. DA; Last. JA; Mellick. PW: Parks. WC: Pinkerton. KE;
Radhakrishnamurthv. B; Ryan. LM; Szarek. JL. (1995a). Consequences of prolonged inhalation of ozone
on F344/N rats: Collaborative studies Part XI: Integrative summary. Cambridge, MA: Health Effects
Institute.
Catalano. PJ: Rogus. J: Ryan. LM. (1995b). Consequences of prolonged inhalation of ozone on F344/N rats:
collaborative studies Part X: robust composite scores based on median polish analysis. Health Effects
Institute.
Chang. L; Miller. FJ; Ultman. J: Huang. Y; Stockstill. BL; Grose. E; Graham. JA; Ospital. JJ: Crapo. JD.
(1991). Alveolar epithelial cell injuries by subchronic exposure to low concentrations of ozone correlate
with cumulative exposure. Toxicol Appl Pharmacol 109: 219-234. http://dx.doi.org/10.1016/0041-
008X(91)90170-J
Chang. LY; Huang. Y; Stockstill. BL; Graham. JA; Grose. EC: Menache. MG: Miller. FJ: Costa. PL; Crapo.
JD. (1992). Epithelial injury and interstitial fibrosis in the proximal alveolar regions of rats chronically
exposed to a simulated pattern of urban ambient ozone. Toxicol Appl Pharmacol 115: 241-252.
http://dx.doi.org/10.1016/0041-008X(92)90329-Q
Chang. LY; Stockstill. BL: Menache. MG: Mercer. RR; Crapo. JD. (1995). Consequences of prolonged
inhalation of ozone on F344/N rats: Collaborative studies. Part VIII: Morphometric analysis of structural
alterations in alveolar regions [HEI] (pp. 3-39). (Research Report 65-VIII). Health Effects Institute.
http://pubs.healtheffects.org/view.php?id=82
Charpin. D; Pascal. L; Birnbaum. J: Armengaud. A; Sambuc. R; Lanteaume. A; Vervloet. D. (1999). Gaseous
air pollution and atopy. Clin Exp Allergy 29: 1474-1480.
7-93
-------
Chen. C: Arjomandi. M; Balmes. J: Tager. I; N. H. (2007a). Effects of chronic and acute ozone exposure on
lipid peroxidation and antioxidant capacity in healthy young adults. Environ Health Perspect 115: 1732-
1737. http://dx.doi.org/10.1289/ehp.10294
Chen. C: Arjomandi. M; Qin. H; Balmes. J: Tager. I; Holland. N. (2006a). Cytogenetic damage in buccal
epithelia and peripheral lymphocytes of young healthy individuals exposed to ozone. Mutagenesis 21: 131-
137. http://dx.doi.org/10.1093/mutage/gel007
Chen. JC: Schwartz. J. (2009). Neurobehavioral effects of ambient air pollution on cognitive performance in
US adults. Neurotoxicology 30: 231-239. http://dx.doi.Org/10.1016/i.neuro.2008.12.011
Chen. L; Bell. EM; Caton. AR; Druschel. CM; Lin. S. (2010a). Residential mobility during pregnancy and the
potential for ambient air pollution exposure misclassification. Environ Res 110: 162-168.
http://dx.doi.0rg/10.1016/i.envres.2009.ll.001
Chen. L; Yang. W: Jennison. BL; Goodrich. A; Omave. ST. (2002). Air pollution and birth weight in northern
Nevada, 1991-1999. Inhal Toxicol 14: 141-157.
Chen. LH; Knutsen. SF; Shavlik. D; Beeson. WL; Petersen. F; Ghamsary. M; Abbey. D. (2005).
The association between fatal coronary heart disease and ambient paniculate air pollution: Are females at
greater risk? Environ Health Perspect 113: 1723-1729.
Chuang. GC: Yang. Z; Westbrook. DG: Pompilius. M; Ballinger. CA; White. RC: Krzywanski. DM;
Postlethwait. EM; Ballinger. SW. (2009). Pulmonary ozone exposure induces vascular dysfunction,
mitochondria! damage, and atherogenesis. Am J Physiol Lung Cell Mol Physiol 297: L209-L216.
httrj://dx.doi.org/10.1152/aiplung.00102.2009
Chuang. KJ; Yan. YH; Chiu. SY; Cheng. TJ. (2011). Long-term air pollution exposure and risk factors for
cardiovascular diseases among the elderly in Taiwan. Occup Environ Med 68: 64-68.
http://dx.doi.org/10.1136/oem.2009.052704
Clark. NA; Demers. PA; Karr. CJ: Koehoorn. M; Lencar. C: Tamburic. L; Brauer. M. (2010). Effect of early
life exposure to air pollution on development of childhood asthma. Environ Health Perspect 118: 284-290.
http://dx.doi.org/10.1289/ehp.0900916
Cole. MP; Freeman. BA. (2009). Promotion of cardiovascular disease by exposure to the air pollutant ozone
[Review]. Am J Physiol Lung Cell Mol Physiol 297: L209-L216.
http://dx.doi.org/10.1152/ajplung.00187.2009
Conti. DV: Cortessis. V: Molitor. J: Thomas. DC. (2003). Bayesian modeling of complex metabolic pathways.
Hum Hered 56: 83-93. http://dx.doi.org/10.1159/000073736
Dadvand. P; Rankin. J: Rushton. S: Pless-Mulloli. T. (2011). Ambient air pollution and congenital heart
disease: A register-based study. Environ Res 111: 435-441. http://dx.doi.Org/10.1016/i.envres.2011.01.022
Dales. R: Burnett. RT; Smith-Doiron. M; Stieb. DM; Brook. JR. (2004). Air pollution and sudden infant death
syndrome. Pediatrics 113: 628-631.
Darrow. LA; Klein. M; Flanders. WD; Waller. LA; Correa. A; Marcus. M; Mulholland. JA; Russell. AG:
Tolbert. PE. (2009). Ambient air pollution and preterm birth: A time-series analysis. Epidemiology 20: 689-
698. http://dx.doi.org/10.1097/EDE.Ob013e3181a7128f
Darrow. LA; Klein. M; Strickland. MJ; Mulholland. JA; Tolbert. PE. (20 lib). Ambient air pollution and birth
weight in full-term infants in Atlanta, 1994-2004. Environ Health Perspect 119: 731-737.
http://dx.doi.org/10.1289/ehp.1002785
DeH'Omo. G: Wolfer. D; Alleva. E; Lipp. HP. (1995). Developmental exposure to ozone induces subtle
changes in swimming navigation of adult mice. Toxicol Lett 81: 91-99.
Diaz. J; Linares. C: Garcia-Herrera. R; Lopez. C: Trigo. R. (2004). Impact of temperature and air pollution on
the mortality of children in Madrid. J Occup Environ Med 46: 768-774.
7-94
-------
Dockery. DW: Pope. CA. Ill; Xu. X; Spengler. JD; Ware. JH; Fay. ME; Ferris. BG. Jr; Speizer. FE. (1993).
An association between air pollution and mortality in six US cities. N Engl J Med 329: 1753-1759.
http://dx.doi.org/10.1056/NEJM199312093292401
Dugandzic. R; Dodds. L; Stieb. D; Smith-Doiron. M. (2006). The association between low level exposures to
ambient air pollution and term low birth weight: A retrospective cohort study. Environ Health 5:3.
http://dx.doi.org/10.1186/1476-069X-5-3
Ercan. H; Birben. E; Dizdar. EA; Keskin. O; Karaaslan. C: Sover. OU; Put. R; Sackesen. C: Besler. T;
Kalavci. O. (2006). Oxidative stress and genetic and epidemiologic determinants of oxidant injury in
childhood asthma. J Allergy Clin Immunol 118: 1097-1104. http://dx.doi.Org/10.1016/j.jaci.2006.08.012
Eustis. SL; Schwartz. LW: Kosch. PC: Dungworth. PL. (1981). Chronic bronchiolitis in nonhuman primates
after prolonged ozone exposure. Am J Pathol 105: 121-137.
Evans. MJ; Fanucchi. MV; Baker. GL; Van Winkle. LS: Pantle. LM; Nishio. SJ: Schelegle. ES: Gershwhin. LJ;
Miller. LA; Hyde. DM; Sannes. PL; Plopper. CG. (2003). Atypical development of the tracheal basement
membrane zone of infant rhesus monkeys exposed to ozone and allergen. Am J Physiol 285: L931-L939.
http://dx.doi.org/10.1152/ajplung.00175.2003
Evans. MJ: Fanucchi. MV: Baker. GL; Van Winkle. LS: Pantle. LM; Nishio. SJ: Schelegle. ES: Gershwin. LJ:
Miller. LA; Hyde. DM; Plopper. CG. (2004). The remodelled tracheal basement membrane zone of infant
rhesus monkeys after 6 months of recovery. Clin Exp Allergy 34: 1131-1136.
http://dx.doi.0rg/10.llll/i.1365-2222.2004.02004.x CEA2004
Exner. M; Minar. E; Wagner. O: Schillinger. M. (2004). The role of heme oxygenase-1 promoter
polymorphisms in human disease [Review]. Free Radic Biol Med 37: 1097-1104.
http://dx.doi.0rg/10.1016/i.freeradbiomed.2004.07.008
Fanucchi. MV: Plopper. CG: Evans. MJ: Hyde. DM; Van Winkle. LS: Gershwin. LJ: Schelegle. ES. (2006).
Cyclic exposure to ozone alters distal airway development in infant rhesus monkeys. Am J Physiol Lung
Cell Mol Physiol 291: L644-L650. http://dx.doi.org/10.1152/aiplung.00027.2006
Fanucchi. MV: Wong. VJ; Hinds. D; Tarkington. BK; Van Winkle. LS: Evans. MJ: Plopper. CG. (2000).
Repeated episodes of exposure to ozone alters postnatal development of distal conducting airways in infant
rhesus monkeys [Abstract]. Am J Respir Crit Care Med 161: A615.
Feichtinger. W: Papalambrou. K; Poehl. M; Krischker. U; Neumann. K. (1997). Smoking and in vitro
fertilization: Ameta-analysis. J Assist Reprod Genet 14: 596-599.
http://dx.doi.Org/10.1023/A:1022584802711
Ferng. SF; Castro. CE; AFifi. AA; Bermudez. E; Mustafa. MG. (1997). Ozone-induced DNA strand breaks in
guinea pig tracheobronchial epithelial cells. J Toxicol Environ Health 51: 353-367.
Forbes. LJ: Patel. MD; Rudnicka. AR; Cook. DG: Bush. T; Stedman. JR; Whincup. PH; Strachan. DP;
Anderson. RH. (2009a). Chronic exposure to outdoor air pollution and markers of systemic inflammation.
Epidemiology 20: 245-253. http://dx.doi.org/10.1097/EDE.Ob013e318190ea3f
Forbes. LJL; Kapetanakis. V: Rudnicka. AR; Cook. DG: Bush. T; Stedman. JR; Whincup. PH; Strachan. DP;
Anderson. HR. (2009b). Chronic exposure to outdoor air pollution and lung function in adults. Thorax 64:
657-663. http://dx.doi.org/10.1136/thx.2008.109389
Frischer. T; Studnicka. M; Gartner. C: Tauber. E; Horak. F; Veiter. A; Spengler. J: Kuhr. J: Urbanek. R. (1999).
Lung function growth and ambient ozone: A three-year population study in school children. Am J Respir
Crit Care Med 160: 390-396.
Frischer. T; Studnicka. M; Halmerbauer. G: Horak. F; Gartner. C: Tauber. E; Koller. DY. (2001). Ambient
ozone exposure is associated with eosinophil activation in healthy children. Clin Exp Allergy 31: 1213-
1219.
Fujinaka. LE; Hyde. DM; Plopper. CG: Tyler. WS: Dungworth. PL; Lollini. LO. (1985). Respiratory
bronchiolitis following long-term ozone exposure in bonnet monkeys: A morphometric study. Exp Lung
Res 8: 167-190.
7-95
-------
Gauderman. WJ: Avol. E; Gilliland. F; Vora. H; Thomas. D; Berhane. K; McConnell. R; Kuenzli. N; Lurmann.
F; Rappaport. E; Margolis. H; Bates. D; Peters. J. (2004). The effect of air pollution on lung development
from 10 to 18 years of age. N Engl J Med 351: 1057-1067.
Gauderman. WJ: McConnell. R; Gilliland. F; London. S: Thomas. D: Avol. E; Vora. H; Berhane. K;
Rappaport. EB; Lurmann. F; Margolis. HG: Peters. J. (2000). Association between air pollution and lung
function growth in southern California children. Am J Respir Crit Care Med 162: 1383-1390.
Gilboa. SM; Mendola. P; Olshan. AF; Langlois. PH; Savitz. DA; Loomis. D: Herring. AH; Fixler. DE. (2005).
Relation between ambient air quality and selected birth defects, seven county study, Texas, 1997-2000. Am
J Epidemiol 162: 238-252. http://dx.doi.org/10.1093/aje/kwil89
Gilliland. FD; Berhane. K; Rappaport. EB; Thomas. DC: Avol. E; Gauderman. WJ: London. SJ: Margolis. HG:
McConnell. R: Islam. KT; Peters. JM. (2001). The effects of ambient air pollution on school absenteeism
due to respiratory illnesses. Epidemiology 12: 43-54.
Gilliland. FD; McConnell. R: Peters. J: Gong. J. r. H. (1999). A theoretical basis for investigating ambient air
pollution and children's respiratory health [Review]. Environ Health Perspect 107: 403-407.
Gilliland. FD: Rappaport. EB: Berhane. K; Islam. T; Dubeau. L; Gauderman. WJ: McConnell. R. (2002).
Effects of glutathione S-Transferase PI, Ml, and Tl on acute respiratory illness in school children. Am J
Respir Crit Care Med 166: 346-351.
Giovannelli. L; Pitozzi. V; Moretti. S: Boddi. V: Dolara. P. (2006). Seasonal variations of DNA damage in
human lymphocytes: Correlation with different environmental variables. Mutat Res-Fundam Mol Mech
Mutagen593: 143-152. http://dx.doi.Org/10.1016/i.mrfmmm.2005.07.002
Gonzalez-Pina. R: Escalante-Membrillo. C: Alfaro-Rodriguez. A; Gonzalez-Maciel. A. (2008). Prenatal
exposure to ozone disrupts cerebellar monoamine contents in newborn rats. Neurochem Res 33: 912-918.
http://dx.doi.org/10.1007/sll064-007-9534-3
Goss. CH; Newsom. SA; Schildcrout JS: Sheppard. L; Kaufman. JD. (2004). Effect of ambient air pollution
on pulmonary exacerbations and lung function in cystic fibrosis. Am J Respir Crit Care Med 169: 816-821.
Gouveia. N; Bremner. SA; Novaes. HMD. (2004). Association between ambient air pollution and birth weight
in Sao Paulo, Brazil. J Epidemiol Community Health 58: 11-17.
Greer. JR; Abbey. DE; Burchette. RJ. (1993). Asthma related to occupational and ambient air pollutants in
nonsmokers. J Occup Environ Med 35: 909-915.
Guerrero. AL; Dorado-Martinez. C: Rodriguez. A; Pedroza-Rios. K; Borgonio-Perez. G: Rivas-Arancibia. S.
(1999). Effects of vitamin E on ozone-induced memory deficits and lipid peroxidation in rats. Neuroreport
10: 1689-1692.
Guevara-Guzman. R: Arriaga. V: Kendrick. KM; Bernal. C: Vega. X; Mercado-Gomez. OF; Rivas-Arancibia.
S.. (2009). Estradiol prevents ozone-induced increases in brain lipid peroxidation and impaired social
recognition memory in female rats. Neuroscience 159: 940-950.
http://dx.doi.0rg/10.1016/i.neuroscience.2009.01.047
Ha. EH; Hong. YC: Lee. BE; Woo. BH; Schwartz. J: Christiani. DC. (2001). Is air pollution a risk factor for
low birth weight in Seoul? Epidemiology 12: 643-648.
Ha. EH; Lee. JT; Kim. H; Hong. YC: Lee. (2003). Infant susceptibility of mortality to air pollution in Seoul,
South Korea. Pediatrics 111: 284-290.
Hack. M; Fanaroff. AA. (1999). Outcomes of children of extremely low birth weight and gestational age in the
1990s [Review]. Early Hum Dev 53: 193-218. http://dx.doi.org/10.1016/S0378-3782(98)00052-8
Hajat. S: Armstrong. B; Wilkinson. P; Busby. A; Polk. H. (2007). Outdoor air pollution and infant mortality:
Analysis of daily time-series data in 10 English cities. J Epidemiol Community Health 61: 719-722.
http://dx.doi.org/10.1136/iech.2006.053942
7-96
-------
Han. SG: Bhoopalan. V; Akinbiyi. T; Gairola. CG: Bhalla. DK. (2011). In utero tobacco smoke exposure alters
pulmonary responses of newborn rats to ozone. J Toxicol Environ Health A 74: 668-677.
http://dx.doi.org/10.1080/15287394.2011.539133
Hanene. C: Jihene. L; Jame. A; Kamel. H; Agnes. H. (2007). Association of GST genes polymorphisms with
asthma in Tunisian children. Mediators Inflamm 19564: 6. http://dx.doi.org/10.1155/2007/19564
Hansen. C: Luben. TJ; Sacks. JD; Olshan. A; Jeffav. S: Strader. L; Perreault. SD. (2010). The effect of ambient
air pollution on sperm quality. Environ Health Perspect 118: 203-209.
http://dx.doi.org/10.1289/ehp.0901022
Hansen. C: Neller. A; Williams. G: Simpson. R. (2006). Maternal exposure to low levels of ambient air
pollution and preterm birth in Brisbane, Australia. BJOG 113: 935-941. http://dx.doi.org/10.llll/U471-
0528.2006.01010.x
Hansen. C: Neller. A; Williams. G: Simpson. R. (2007). Low levels of ambient air pollution during pregnancy
and fetal growth among term neonates in Brisbane, Australia. Environ Res 103: 383-389.
http://dx.doi.0rg/10.1016/i.envres.2006.06.010
Hansen. CA; Barnett. AG: Jalaludin. B; Morgan. G. (2009). Ambient air pollution and birth defects in
Brisbane, Australia. PLoS ONE 4: e5408. http://dx.doi.org/10.1371/iournal.pone.0005408
Hansen. CA; Barnett. AG: Pritchard. G. (2008). The effect of ambient air pollution during early pregnancy on
fetal ultrasonic measurements during mid-pregnancy. Environ Health Perspect 116: 362-369.
http://dx.doi.org/10.1289/ehp.10720
Harkema. JR; Catalano. PJ; Hotchkiss. JA. (1997a). Consequences of prolonged inhalation of ozone on
F344/N rats: Collaborative studies: Part XII. Atrophy of bone in nasal turbinates (pp. 1-19; discussion 21-
16). (ISSN 1041-5505
65-XII). Cambridge, MA: Health Effects Institute.
Harkema. JR: Hotchkiss. JA; Barr. EB; Bennett. CB; Gallup. M; Lee. JK; Basbaum. C. (1999). Long-lasting
effects of chronic ozone exposure on rat nasal epithelium. Am J Respir Cell Mol Biol 20: 517-529.
Harkema. JR: Hotchkiss. JA: Griffith. WC. (1997b). Mucous cell metaplasia in rat nasal epithelium after a 20-
month exposure to ozone: A morphometric study of epithelial differentiation. Am J Respir Cell Mol Biol
16: 521-530.
Harkema. JR: Morgan. KT; Gross. EA; Catalano. PJ: Griffith. WC. (1994). Consequences of prolonged
inhalation of ozone on F344/N rats: Collaborative studies Part VII: Effects on the nasal mucociliary
apparatus. Cambridge, MA: Health Effects Institute.
Harkema. JR: Plopper. CG: Hyde. DM; St George. JA: Dungworth. PL. (1987a). Effects of an ambient level of
ozone on primate nasal epithelial mucosubstances: quantitative histochemistry. Am J Pathol 127: 90-96.
Harkema. JR: Plopper. CG: Hyde. DM; St George. JA: Wilson. DW: Dungworth. PL. (1987b). Response of
the macaque nasal epithelium to ambient levels of ozone: A morphologic and morphometric study of the
transitional and respiratory epithelium. Am J Pathol 128: 29-44.
Harkema. JR: Plopper. CG: Hyde. PM; St George. JA: Wilson. PW: Pungworth. PL. (1993). Response of
macaque bronchiolar epithelium to ambient concentrations of ozone. Am J Pathol 143: 857-866.
Haro. R; Paz. C. (1993). Effects of ozone exposure during pregnancy on ontogeny of sleep in rats. Neurosci
Lett 164: 67-70. http://dx.doi.org/10.1016/0304-3940(93)90859-J
Hassett. C: Mustafa. MG: Coulson. WF; Elashoff. RM. (1985). Murine lung carcinogenesis following
exposure to ambient ozone concentrations. J Natl Cancer Inst 75: 771-777.
Herbert. RA; Hailev. JR: Grumbein. S: Chou. BJ; Sills. RC: Haseman. JK; Goehl. T; Miller. RA; Roycroft. JH;
Boorman. GA. (1996). Two-year and lifetime toxicity and carcinogenicity studies of ozone in B6C3F1
mice. Toxicol Pathol 24: 539-548.
7-97
-------
Himes. BE; Hunninghake. GM; Baurlev. JW: Rafaels. NM; Sleiman. P; Strachan. DP; Wilk. JB; Willis-Owen.
SAG: Klanderman. B; Lasky-Su. J: Lazarus. R; Murphy. AJ; Soto-Quiros. ME; Avila. L; Beatv. T; Mathias.
RA; Ruczinski. I; Barnes. KG: Celedon. JC: Cookson. WOC: Gauderman. WJ: Gilliland. FD; Hakonarson.
H; Lange. C: Moffatt. MF; O'Connor. GT; Rabv. BA; Silverman. EK; Weiss. ST. (2009). Genome-wide
Association Analysis Identifies PDE4D as an Asthma-Susceptibility Gene. Am J Hum Genet 84: 581-593.
http://dx.doi.0rg/10.1016/i.aihg.2009.04.006
Hoffjan. S: Nicolae. D; Ostrovnava. I; Roberg. K; Evans. M; Mirel. DB; Steiner. L; Walker. K; Shult. P;
Gangnon. RE; Gern. JE; Martinez. FD; Lemanske. RF; Ober. C. (2005). Gene-environment interaction
effects on the development of immune responses in the 1st year of life. Am J Hum Genet 76: 696-704.
http://dx.doi.org/10.1086/429418
Holgate. ST; Davies. DE; Powell. PJVI; Howarth. PH; Haitchi. HM; Hollowav. JW. (2007). Local genetic and
environmental factors in asthma disease pathogenesis: chronicity and persistence mechanisms [Review].
Eur Respir J 29: 793-803. http://dx.doi.org/10.1183/09031936.00087506
Horak. F. Jr; Studnicka. M; Gartner. C: Spengler. JD; Tauber. E; Urbanek. R; Veiter. A; Frischer. T. (2002).
Paniculate matter and lung function growth in children: a 3-yr follow-up study in Austrian schoolchildren.
Eur Respir J 19: 838-845.
Huen. K; Gunn. L; Duramad. P: Jeng. M; Scalf. R: Holland. N. (2006). Application of a geographic
information system to explore associations between air pollution and micronucleus frequencies in African
American children and adults. Environ Mol Mutagen 47: 236-246. http://dx.doi.org/10.1002/em.20193
Hunter. DJ. (2005). Gene-environment interactions in human diseases [Review]. Nat Rev Genet 6: 287-298.
http://dx.doi.org/10.1038/nrgl578
Hutcheon. JA; Platt. RW (2008). The missing data problem in birth weight percentiles and thresholds for
"small-for-gestational-age". Am J Epidemiol 167: 786-792. http://dx.doi.org/10.1093/aje/kwm327
Hwang. BF; Jaakkola. JJ. (2008). Ozone and other air pollutants and the risk of oral clefts. Environ Health
Perspect 116: 1411-1415. http://dx.doi.org/10.1289/ehp.11311
Hwang. BF; Jaakkola. JJK; Lee. YL; Lin. YC: Y-LL. G. (2006). Relation between air pollution and allergic
rhinitis in Taiwanese schoolchildren. Respir Res 7: 23.
Hwang. BF; Lee. YL; Lin. YC: Jaakkola. JJK: Guo. YL. (2005). Traffic related air pollution as a determinant
of asthma among Taiwanese school children. Thorax 60: 467-473.
Hyde. DM; Hubbard. WC: Wong. V; Wu. R: Pinkerton. K; Plopper. CG. (1992). Ozone-induced acute
tracheobronchial epithelial injury: relationship to granulocyte emigration in the lung. Am J Respir Cell Mol
Biol 6: 481-497.
Hyde. DM; Miller. LA; Schelegle. ES: Fanucchi. MV; Van Winkle. LS: Tyler. NK; Avdalovic. MV; Evans. MJ;
Kaiekar. R; Buckpitt. AR; Pinkerton. KE; Joad. JP; Gershwin. LJ; Wu. R: Plopper. CG. (2006). Asthma: a
comparison of animal models using stereological methods. Eur Respir Rev 15: 122-135.
http://dx.doi.org/10.1183/09059180.00010103
Hyde. DM; Plopper. CG: Harkema. JR: St George. JA; Tyler. WS: Dungworth. PL. (1989). Ozone-induced
structural changes in monkey respiratory system. In T Schneider; SD Lee; GJR Wolters; LD Grant (Eds.),
Atmospheric ozone research and its policy implications: Proceedings of the 3rd US-Dutch International
Symposium, Nijmegen, the Netherlands May 9-13, 1988 (pp. 523-532). Amsterdam, The Netherlands:
Elsevier Science Publishers.
Ihorst. G; Frischer. T; Horak. F; Schumacher. M; Kopp. M; Forster. J; Mattes. J; Kuehr. J. (2004). Long- and
medium-term ozone effects on lung growth including a broad spectrum of exposure. Eur Respir J 23: 292-
299.
lijima. MK; Kobavashi. T. (2004). Nasal allergy-like symptoms aggravated by ozone exposure in a
concentration-dependent manner in guinea pigs. Toxicology 199: 73-83.
http://dx.doi.0rg/10.1016/i.tox.2004.01.008
7-98
-------
Islam. T; Berhane. K; Mcconnell. R; Gauderman. WJ: Avol. E; Peters. JM; Gilliland. FD. (2009). Glutathione-
S-transferase (GST) PI, GSTM1, exercise, ozone and asthma incidence in school children. Thorax 64: 197-
202. http://dx.doi.org/10.1136/tfax.2008.099366
Islam. T; Gauderman. WJ: Berhane. K; Mcconnell. R; Avol. E; Peters. JM; Gilliland. FD. (2007).
The relationship between air pollution, lung function and asthma in adolescents. Thorax 62: 957-963.
http://dx.doi.org/10.1136/thx.2007.078964
Islam. T; Mcconnell. R: Gauderman. WJ: Avol. E; Peters. JM: Gilliland. FD. (2008). Ozone, oxidant defense
genes and risk of asthma during adolescence. Am J Respir Crit Care Med 177: 388-395.
http://dx.doi.org/10.1164/rccm.200706-863OC
Jacquemin. B; Kauffmann. F; Pin. I; Le Moual. N; Bousquet. J: Gormand. F; Just. J: Nadif. R: Pison. C:
Vervloet. D; Kiinzli. N; Siroux. V. (In Press) Air pollution and asthma control in the epidemiological study
on the genetics and environment of asthma. J Epidemiol Community Health.
http://dx.doi.org/10.1136/iech.2010.130229
Jakab. GJ: Bassett. DJP (1990). Influenza virus infection, ozone exposure, and fibrogenesis. Am J Respir Crit
Care Med 141: 1307-1315.
Jalaludin. B; Mamies. T; Morgan. G: Lincoln. D; Sheppeard. V: Corbett. S. (2007). Impact of ambient air
pollution on gestational age is modified by season in Sydney, Australia. Environ Health 6: 16.
http://dx.doi.org/10.1186/1476-069X-6-16
Jedlinska-Krakowska. M; Bomba. G: Jakubowski. K; Rotkiewicz. T; Jana. B; Penkowskii. A. (2006). Impact
of oxidative stress and supplementation with vitamins E and C on testes morphology in rats. J Reprod Dev
52: 203-209.
Jerrett. M; Burnett. RT; Pope. CA. Ill; Ito. K; Thurston. G: Krewski. D; Shi. Y; Calle. E; Thun. M. (2009).
Long-term ozone exposure and mortality. N Engl J Med 360: 1085-1095.
http://dx.doi.org/10.1056/NEJMoa0803894
Jiang. LL; Zhang. YH; Song. GX; Chen. GH; Chen. BH; Zhao. NO; Kan. HP. (2007). A time series analysis of
outdoor air pollution and preterm birth in Shanghai, China. Biomed Environ Sci 20: 426-431.
Joad. JP: Kott. KS: Brie. JM: Peake. JL; Plopper. CG: Schelegle. ES: Gershwin. LJ; Pinkerton. KE. (2006).
Structural and functional localization of airway effects from episodic exposure of infant monkeys to
allergen and/or ozone. Toxicol Appl Pharmacol 214: 237-243. http://dx.doi.Org/10.1016/i.taap.2005.12.012
Joad. JP: Kott. KS: Brie. JM: Schelegle. ES: Gershwin. LJ: Plopper. CG: Peake. JL; Pinkerton. KE. (2008).
The effects of inhaled corticosteroids on intrinsic responsiveness and histology of airways from infant
monkeys exposed to house dust mite allergen and ozone. Toxicol Appl Pharmacol 226: 153-160.
http://dx.doi.0rg/10.1016/i.taap.2007.09.005
Kajekar. R; Pieczarka. EM; Smiley-Jewell. SM; Schelegle. ES: Fanucchi. MV; Plopper. CG. (2007). Early
postnatal exposure to allergen and ozone leads to hyperinnervation of the pulmonary epithelium. Respir
Physiol Neurobiol 155: 55-63. http://dx.doi.Org/10.1016/i.resp.2006.03.002
Karr. C: Lumlev. T; Schreuder. A; Davis. R: Larson. T; Ritz. B; Kaufman. J. (2007). Effects of subchronic and
chronic exposure to ambient air pollutants on infant bronchiolitis. Am J Epidemiol 165: 553-560.
http://dx.doi.org/10.1093/aie/kwk032
Katre. A; Ballinger. C: Akhter. H; Fanucchi. M; Kim. DK; Postlethwait E; Liu. RM. (2011). Increased
transforming growth factor beta 1 expression mediates ozone-induced airway fibrosis in mice. Inhal
Toxicol 23: 486-494. http://dx.doi.org/10.3109/08958378.2011.584919
Kavlock. R; Daston. G: Grabowski. CT. (1979). Studies on the developmental toxicity of ozone. I. Prenatal
effects. Toxicol Appl Pharmacol 48: 19-28. http://dx.doi.org/10.1016/S0041-008X(79)80004-6
Kavlock. RJ; Meyer. E; Grabowski. CT. (1980). Studies on the developmental toxicity of ozone: Postnatal
effects. Toxicol Lett 5: 3-9. http://dx.doi.org/10.1016/0378-4274(80)90141-l
7-99
-------
Kim. MY; Cho. MY. (2009a). Toxicity and carcinogenicity of ozone in combination with 4-(N-methyl-N-
nitrosamino)-l-(3-pyridyl)-l-butanone and dibutyl phthalate in B6C3F1 mice for 16 and 32 weeks. Biomed
Environ Sci 22: 216-222. http://dx.doi.org/10.1016/S0895-3988(09)60048-9
Kim. MY; Cho. MY. (2009b). Tumorigenesis in B6C3F1 mice exposed to ozone in combination with 4-(N-
methyl-N-nitrosamino)-l-(3-pyridyl)-l-butanone and dietary dibutyl phthalate. Toxicol Ind Health 25: 189-
195. http://dx.doi.org/10.1177/0748233709106185
Kinnev. PL; Lippmann. M. (2000). Respiratory effects of seasonal exposures to ozone and particles. Arch
Environ Occup Health 55: 210-216.
Kleeberger. SR; Peden. D. (2005). Gene-environment interactions in asthma and other respiratory diseases
[Review]. Annu Rev Med 56: 383-400. http://dx.doi.org/10.1146/annurev.med.56.062904.144908
Kodavanti. UP; Thomas. R; Ledbetter. AD; Schladweiler. MC: Shannahan. JH; Wallenborn. JG: Lund. AK;
Campen. MJ; Butler. EO; Gottipolu. RR; Nyska. A; Richards. JE; Andrews. D; Jaskot. RH; Mckee. J:
Kotha. S. R.: Patel. RB; Parianandi. NL. (2011). Vascular and cardiac impairments in rats Inhaling ozone
and diesel exhaust particles. Environ Health Perspect 119: 312-318. http://dx.doi.org/10.1289/ehp.1002386
Krewski. D; Jerrett. M; Burnett. RT; Ma. R: Hughes. E; Shi. Y; Turner. MC: 3rd. PA; Thurston. G: Calle. EE;
Thun. MJ; Beckerman. B; DeLuca. P; Finkelstein. N; Ito. K; Moore. DK; Newbold. KB; Ramsay. T; Ross.
Z; Shin. H; Tempalski. B. (2009). Extended follow-up and spatial analysis of the American Cancer Society
study linking paniculate air pollution and mortality (pp. 5-114; discussion 115-136). (ISSN 1041-5505
Kiinzli. N. (2012). Is air pollution of the 20th century a cause of current asthma hospitalisations? [Editorial].
Thorax 67: 2-3. http://dx.doi.org/10.1136/thoraxinl-2011-200919
Kuo. HW: Lai. JS: Lee. MC: Tai. RC: Lee. MC. (2002). Respiratory effects of air pollutants among asthmatics
in central Taiwan. Arch Environ Occup Health 57: 194-200.
Lampl. M; Jeantv. R (2003). Timing is everything: A reconsideration of fetal growth velocity patterns identifies
the importance of individual and sex differences. Am J Hum Biol 15: 667-680.
http://dx.doi.org/10.1002/aihb.10204
Larson. SD; Schelegle. ES: Walbv. WF; Gershwin. LJ; Fanuccihi. MV; Evans. MJ: Joad. JP; Tarkington. BK;
Hyde. DM; Plopper. CG. (2004). Postnatal remodeling of the neural components of the epithelial-
mesenchymal trophic unit in the proximal airways of infant rhesus monkeys exposed to ozone and allergen.
Toxicol Appl Pharmacol 194: 211-220. http://dx.doi.Org/10.1016/j.taap.2003.09.025
Last. JA; Gelzleichter. TR; Harkema. J: Hawk. S. (1994). Consequences of prolonged inhalation of ozone on
Fischer-344/N rats: Collaborative studies. Part I: Content and cross-linking of lung collagen [HEI].
(Research Report 65-1). Cambridge, MA: Health Effects Institute.
http://pubs.healtheffects.org/view.php?id=81
Last. JA; Reiser. KM; Tyler. WS: Rucker. RB. (1984). Long-term consequences of exposure to ozone. I. Lung
collagen content. Toxicol Appl Pharmacol 72: 111-118.
Last. JA; Warren. PL; Pecquet-Goad. E; Witschi. H. (1987). Modification by ozone of lung tumor
development in mice. J Natl Cancer Inst 78: 149-154.
Latzin. P; Roosli. M; Huss. A; Kuehni. CE; Frev. U. (2009). Air pollution during pregnancy and lung function
in newborns: Abirth cohort study. Eur Respir J 33: 594-603. http://dx.doi.org/10.1183/09031936.00084008
Lee. SJ: Hajat. S: Steer. PJ; Filippi. V. (2008c). A time-series analysis of any short-term effects of
meteorological and air pollution factors on preterm births in London, UK. Environ Res 106: 185-194.
Lee. YL; Lin. YC: Lee. YC: Wang. JY; Hsiue. TR; Guo. YL. (2004b). Glutathione S-transferase PI gene
polymorphism and air pollution as interactive risk factors for childhood asthma. Clin Exp Allergy 34: 1707-
1713.
Lee. YL; Mcconnell. R: Berhane. K; Gilliland. FD. (2009b). Ambient ozone modifies the effect of tumor
necrosis factor G-308A on bronchitic symptoms among children with asthma. Allergy 64: 1342-1348.
http://dx.doi.0rg/10.llll/i.1398-9995.2009.02014.x
7-100
-------
Legro. RS: Sauer. MV; Mottla. GL; Richter. KS: Li. X; Dodson. WC: Liao. D. (2010). Effect of air quality on
assisted human reproduction. Hum Reprod 25: 1317-1324. http://dx.doi.org/10.1093/humrep/deq021
Li. H; Romieu. I; Sienra-Monge. JJ: Ramirez-Aguilar. M; Estela del Rio-Navarro. B; Kistner. EO; Giessing.
HK; Irma del Carmen. LS: Chiu. GY; London. SJ. (2006a). Genetic polymorphisms in arginase I and II and
childhood asthma and atopy. J Allergy Clin Immunol 117: 119126.
Li. YF; Gauderman. WJ: Conti. DV: Lin. PC: Avol. E; Gilliland. FD. (2008). Glutathione S-Transferase PI,
Maternal Smoking, and Asthma in Children: A Haplotype-Based Analysis. Environ Health Perspect 116:
409-415.
Lin. CA; Pereira. LAA; Nishioka. DC: Conceicao. GMS: Graga. ALF; Saldiva. PHN. (2004a). Air pollution
and neonatal deaths in Sao Paulo, Brazil. Braz J Med Biol Res 37: 765-770.
Lin. CM; Li. CY; Yang. GY; Mao. IF. (2004b). Association between maternal exposure to elevated ambient
sulfur dioxide during pregnancy and term low birth weight. Environ Res 96: 41-50.
http://dx.doi.0rg/10.1016/i.envres.2004.03.005
Lin. S. (2010). E-mail correspondence from Shao Lin to Dennis Kotchmar dated December 20, 2010.
Available online
Lin. S: Fitzgerald. E; Hwang. SA; Munsie. JP; Stark. A. (1999). Asthma hospitalization rates and
socioeconomic status in New York State (1987-1993). J Asthma 36: 239-251.
Lin. S: Liu. X; Le. LH; Hwang. SA. (2008b). Chronic exposure to ambient ozone and asthma hospital
admissions among children. Environ Health Perspect 116: 1725-1730. http://dx.doi.org/10.1289/ehp.11184
Linn. WS: Rappaport EB; Berhane. KT; Bastain. TM; Avol. EL; Gilliland. FD. (2009). Exhaled nitric oxide in
a population-based study of southern California schoolchildren. Respir Res 10: 28.
http://dx.doi.org/10.1186/1465-9921-10-28
Lipfert. FW: Batv. JD; Miller. JP; Wyzga. RE. (2006a). PM2.5 constituents and related air quality variables as
predictors of survival in a cohort of U.S. military veterans. Inhal Toxicol 18: 645-657.
http://dx.doi.org/10.1080/08958370600742946
Lipfert. FW: Perry. HM. Jr; Miller. JP: Batv. JD: Wyzga. RE; Carmodv. SE. (2000). The Washington
University-EPRI veterans' cohort mortality study: Preliminary results. Inhal Toxicol 4: 41-73.
Lipfert. FW: Perry. HM. Jr; Miller. JP: Batv. JD: Wyzga. RE; Carmodv. SE. (2003). Air pollution, blood
pressure, and their long-term associations with mortality. Inhal Toxicol 15: 493-512.
Lipfert. FW: Wyzga. RE; Batv. JD: Miller. JP. (2006b). Traffic density as a surrogate measure of environmental
exposures in studies of air pollution health effects: Long-term mortality in a cohort of US veterans. Atmos
Environ 40: 154-169.
Liu. S: Krewski. D; Shi. Y; Chen. Y; Burnett. R. (2007b). Association between maternal exposure to ambient
air pollutants during pregnancy and fetal growth restriction. J Expo Sci Environ Epidemiol 17: 426-432.
http://dx.doi.org/10.1038/sj.jes.7500503
Loomis. D: Castillejos. M; Gold. PR; McDonnell. W: Borja-Aburto. VH. (1999). Air pollution and infant
mortality in Mexico City. Epidemiology 10: 118-123.
Lopez. I; Sanchez. I; Bizarro. P; Acevedo. S: Ustarroz. M; Fortoul. T. (2008). Ultrastructural alterations during
embryonic rats' lung development caused by ozone. J Electron Microsc (Tokyo) 57: 19-23.
http://dx.doi.org/10.1093/imicro/dfm033
Maniar-Hew. K; Postlethwait. EM; Fanucchi. MV: Ballinger. CA; Evans. MJ; Harkema. JR: Carey. SA;
McDonald. RJ: Bartolucci. AA; Miller. LA. (2011). Postnatal episodic ozone results in persistent
attenuation of pulmonary and peripheral blood responses to LPS challenge. Am J Physiol Lung Cell Mol
Physiol 300: L462-L471. http://dx.doi.org/10.1152/aiplung.00254.2010
Mamies. T; Jalaludin. B; Morgan. G: Lincoln. D; Sheppeard. V: Corbett. S. (2005). Impact of ambient air
pollution on birth weight in Sydney, Australia. Occup Environ Med 62: 524-530.
http://dx.doi.org/10.1136/oem.2004.014282
7-101
-------
Mariassv. AT; Abraham. WM; Phipps. RJ: Sielczak. MW: Wanner. A. (1990). Effect of ozone on the postnatal
development of lamb mucociliary apparatus. J Appl Physiol 68: 2504-2510.
Mariassv. AT; Sielczak. MW: Mccrav. MN; Abraham. WM; Wanner. A. (1989). Effects of ozone on lamb
trachea! mucosa: Quantitative glycoconjugate histochemistry. Am J Pathol 135: 871-879.
Marshall. E; Harris. G: Wartenberg. D. (2010). Oral cleft defects and maternal exposure to ambient air
pollutants in New Jersey. Birth Defects Res A Clin Mol Teratol 88: 205-215.
http://dx.doi.org/10.1002/bdra.20650
Martinez. FD. (2007a). Gene-environment interactions in asthma: with apologies to William of Ockham
[Review]. Proc Am Thorac Soc 4: 26-31. http://dx.doi.org/10.1513/pats.200607-144JG
Martinez. FD. (2007b). Genes, environments, development and asthma: a reappraisal [Review]. Eur Respir J
29: 179-184. http://dx.doi.org/10.1183/09031936.00087906
Martinez. FD: Wright. AL; Taussig. LM; Holberg. CJ: Halonen. M; Morgan. WJ: Associates. GHM. (1995).
Asthma and wheezing in the first six years of life. N Engl J Med 332: 133-138.
McConnell. R; Berhane. K; Gilliland. F; London. SJ: Islam. T; Gauderman. WJ: Avol. E; Margolis. HG:
Peters. JM. (2002). Asthma in exercising children exposed to ozone: A cohort study. Lancet 359: 386-391.
http://dx.doi.org/10.1016/S0140-6736(02)07597-9
McConnell. R; Berhane. K; Gilliland. F; London. SJ: Vora. H; Avol. E; Gauderman. WJ: Margolis. HG:
Lurmann. F; Thomas. DC: Peters. JM. (1999). Air pollution and bronchitic symptoms in southern
California children with asthma. Environ Health Perspect 107: 757-760.
McConnell. R: Berhane. K; Gilliland. F; Molitor. J: Thomas. D; Lurmann. F; Avol. E; Gauderman. WJ: Peters.
JM. (2003). Prospective study of air pollution and bronchitic symptoms in children with asthma. Am J
Respir Crit Care Med 168: 790-797.
McConnell. R: Islam. T; Shankardass. K; Jerrett. M; Lurmann. F; Gilliland. F; Gauderman. J: Avol. E;
Kuenzli. N; Yao. L; Peters. J: Berhane. K. (2010). Childhood incident asthma and traffic-related air
pollution at home and school. Environ Health Perspect 118: 1021-1026.
httrj://dx.doi.org/10.1289/ehp.09Q1232
McDonnell. WF; Abbey. DE; Nishino. N; Lebowitz. MD. (1999a). Long-term ambient ozone concentration
and the incidence of asthma in nonsmoking adults: the Ahsmog study. Environ Res 80: 110-121.
Meng. YY; Rull. RP; Wilhelm. M; Lombardi. C: Balmes. J: Ritz. B. (2010). Outdoor air pollution and
uncontrolled asthma in the San Joaquin Valley, California. J Epidemiol Community Health 64: 142-147.
http://dx.doi.org/10.1136/iech.2008.083576
Meng. YY; Wilhelm. M; Rull. RP; English. P; Ritz. B. (2007). Traffic and outdoor air pollution levels near
residences and poorly controlled asthma in adults. Ann Allergy Asthma Immunol 98: 455-463.
Miller. LA; Gerriets. JE; Tyler. NK; Abel. K; Schelegle. ES: Plopper. CG: Hyde. DM. (2009). Ozone and
allergen exposure during postnatal development alters the frequency and airway distribution of CD25+
cells in infant rhesus monkeys. Toxicol Appl Pharmacol 236: 39-48.
http://dx.doi.0rg/10.1016/i.taap.2008.12.031
Moffatt. RK; Hyde. DM; Plopper. CG: Tyler. WS: Putney. LF. (1987). Ozone-induced adaptive and reactive
cellular changes in respiratory bronchioles of Bonnet monkeys. Exp Lung Res 12: 57-74.
Monchaux. G: Morlier. JP; Morin. M; Rochefort. P; Maximilien. R; Tredaniel. J. (1996). Co-carcinogenic
effects in rats of combined exposure to radon and ozone. Environ Int 221: S909-S915.
Moore. K; Neugebauer. R; Lurmann. F; Hall. J: Braier. V; Alcorn. S: Tager. I. (2008). Ambient ozone
concentrations cause increased hospitalizations for asthma in children: An 18-year study in Southern
California. Environ Health Perspect 116: 1063-1070. http://dx.doi.org/10.1289/ehp.10497
Morello-Frosch. R; Jesdale. BM; Sadd. JL; Pastor. M. (2010). Ambient air pollution exposure and full-term
birth weight in California. Environ Health 9: 44. http://dx.doi.org/10.1186/1476-069X-9-44
7-102
-------
Morris. CR; Poljakovic. M; Lavrisha. L; Machado. L; Kuypers. FA; Morris. SM. Jr. (2004). Decreased arginine
bioavailability and increased serum arginase activity in asthma. Am J Respir Crit Care Med 170: 148-153.
http://dx.doi.org/10.1164/rccm.200309-1304OC
Mortimer. K; Neugebauer. R; Lurmann. F; Alcorn. S: Balmes. J: Tager. I. (2008a). Air pollution and pulmonary
function in asthmatic children: Effects of prenatal and lifetime exposures. Epidemiology 19: 550-557.
http://dx.doi.org/10.1097/EDE.Ob013e31816a9dcb
Mortimer. K; Neugebauer. R; Lurmann. F; Alcorn. S: Balmes. J: Tager. I. (2008b). Early-lifetime exposure to
air pollution and allergic sensitization in children with asthma. J Asthma 45: 874-881.
http://dx.doi.org/10.1080/02770900802195722
Murphy. SL; Xu. JQ; Kochanek. KD. (2012). Deaths: Preliminary data for 2010. In National Vital Statistics
Reports. (4). Hyattsville, MD: National Center for Health Statistics.
http://www.cdc.gov/nchs/data/nvsr/nvsr60/nvsr60 04.pdf
NHLBI (National Institutes of Health, National Heart Lung and Blood Institute). (2007). Expert panel report 3:
guidelines for the diagnosis and management of asthma. (07-4051). Bethesda, MD: National Institute of
Health.
NTP (National Toxicology Program). (1994). Toxicology and carcinogenesis: Studies of ozone (CAS
No 10028-15-6) and ozone/NNK (CAS No 10028-15-6/64091-91-4) in F344/N rats and B6C3F1 mice (pp.
314). (Technical Report No. 440). Research Triangle Park, NC.
http://ntp.niehs.nih.gov/index.cfm?objectid=070AOEBD-081E-B501-E38F640803C3542C
Ober. C. (2005). Perspectives on the past decade of asthma genetics. J Allergy Clin Immunol 116: 274-278.
http://dx.doi.0rg/10.1016/i.jaci.2005.04.039
Oryszczyn. MP; Bouzigon. E; Maccario. J: Siroux. V; Nadif. R: Wright. A; Kauffmann. F. (2007).
Interrelationships of quantitative asthma-related phenotypes in the epidemiological study on the genetics
and environment of asthma, bronchial hyperresponsiveness, and atopy. J Allergy Clin Immunol 119: 57-63.
Palli. D: Sera. F; Giovannelli. L; Masala. G: Grechi. D; Bendinelli. B; Caini. S: Dolara. P; Saieva. C. (2009).
Environmental ozone exposure and oxidative DNA damage in adult residents of Florence, Italy. Environ
Pollut 157: 1521-1525. http://dx.doi.Org/10.1016/i.envpol.2008.09.011
Parker. JD; Akinbami. LJ; Woodruff. TJ. (2009). Air pollution and childhood respiratory allergies in the United
States. Environ Health Perspect 117: 140-147. http://dx.doi.org/10.1289/ehp.11497
Parker. JD; Rich. DO; Glinianaia. SV: Leem. JH; Wartenberg. D; Bell. ML; Bonzini. M; Brauer. M; Darrow. L;
Gehring. U; Gouveia. N; Grillo. P; Ha. E; van den Hooven. EH; Jalaludin. B; Jesdale. BM; Lepeule. J:
Morello-Frosch. R; Morgan. GG: Slama. R: Pierik. FH; Pesatori. AC: Sathvanaravana. S: Seo. J:
Strickland. M; Tamburic. L; Woodruff. TJ. (2011). The international collaboration on air pollution and
pregnancy outcomes: Initial results. Environ Health Perspect 119: 1023-1028.
http://dx.doi.org/10.1289/ehp.1002725
Paz. C: Bazan-Perkins. B. (1992). Sleep-wake disorganization in cats exposed to ozone. Neurosci Lett 140:
270-272.
Paz. C: Huitron-Resendiz. S. (1996). The effects of ozone exposure on the sleep-wake cycle and serotonin
contents in the pons of the rat. Neurosci Lett 204: 49-52.
Peel. JL; Tolbert. PE; Klein. M; Metzger. KB; Flanders. WD; Knox. T; Mulholland. JA; Ryan. PB; Frumkin.
tL (2005). Ambient air pollution and respiratory emergency department visits. Epidemiology 16: 164-174.
Peluso. M; Munnia. A; Hoek. G: Krzyzanowski. M; Veglia. F; Airoldi. L; Autrup. H; Dunning. A; Garte. S:
Hainaut. P: Malaveille. C: Gormally. E; Matullo. G: Overvad. K; Raaschou-Nielsen. O; Clavel-Chapelon.
F; Linseisen. J: Boeing. H; Trichopoulou. A; Trichopoulos. D: Kaladidi. A; Palli. D: Krogh. V; Tumino. R;
Panico. S: Bueno-De-Mesquita. HB; Peeters. PH; Kumle. M; Gonzalez. CA; Martinez. C: Dorronsoro. M;
Barricarte. A; Navarro. C: Quiros. JR; Berglund. G: Janzon. L; Jarvholm. B; Day. NE; Key. TJ: Saracci. R:
Kaaks. R: Riboli. E; Vineis. P. (2005). DNA adducts and lung cancer risk: A prospective study. Cancer Res
65: 8042-8048. http://dx.doi.org/10.1158/0008-5472.CAN-04-3488
7-103
-------
Penard-Morand. C: Charpin. D; Raherison. C: Kopferschmitt. C: Caillaud. D; Lavaud. F; Annesi-Maesano. I.
(2005). Long-term exposure to background air pollution related to respiratory and allergic health in
schoolchildren. Clin Exp Allergy 35: 1279-1287.
Pereira. FAC: De Assuncao. JV: Saldiva. PHN: Pereira. LAA; Mirra. AP; Braga. ALF. (2005). Influence of air
pollution on the incidence of respiratory tract neoplasm. J Air Waste Manag Assoc 55: 83-87.
Pereira. LAA; Loomis. D: Conceicao. GMS: Braga. ALF: Areas. RM; Kishi. HS: Singer. JM; Bohm. GM;
Saldiva. PHN. (1998). Association between air pollution and intrauterine mortality in Sao Paulo, Brazil.
Environ Health Perspect 106: 325-329.
Perepu. RS: Garcia. C: Postal. D: Sethi. R. (2010). Enhanced death signaling in ozone-exposed ischemic-
reperfused hearts. Mol Cell Biochem 336: 55-64. http://dx.doi.org/10.1007/sll010-009-0265-4
Peters. JM; Avol. E; Gauderman. WJ: Linn. WS: Navidi. W: London. SJ: Margolis. H; Rappaport. E; Vora. H;
Gong. H. Jr; Thomas. DC. (1999a). A study of twelve southern California communities with differing levels
and types of air pollution: II. Effects on pulmonary function. Am J Respir Crit Care Med 159: 768-775.
Peters. JM: Avol. E; Navidi. W: London. SJ: Gauderman. WJ: Lurmann. F; Linn. WS: Margolis. H; Rappaport.
E; Gong. H. Jr; Thomas. DC. (1999b). A study of twelve southern California communities with differing
levels and types of air pollution: I. Prevalence of respiratory morbidity. Am J Respir Crit Care Med 159:
760-767.
Petruzzi. S: De Acetis. L; Chiarotti. F; Sorace. A; Alleva. E. (1999). Limited changes in handedness and
morphine reactivity in CD-I mice after pre- and postnatal ozone exposure. ActaNeurobiol Exp (Wars) 59:
115-122.
Petruzzi. S: Fiore. M; DeH'Omo. G: Bignami. G: Alleva. E. (1995). Medium and long-term behavioral effects
in mice of extended gestational exposure to ozone. Neurotoxicol Teratol 17: 463-470.
Pinkerton. KE; Brodv. AR; Miller. FJ; Crapo. JD. (1989). Exposure to low levels of ozone results in enhanced
pulmonary retention of inhaled asbestos fibers. Am J Respir Crit Care Med 140: 1075-1081.
Pinkerton. KE; Menache. MG: Plopper. CG. (1995). Consequences of prolonged inhalation of ozone on
F344/N rats: Collaborative studies. Part IX: Changes in the tracheobronchial epithelium, pulmonary acinus,
and lung antioxidant enzyme activity [HEI] (pp. 41-98). (Research Report 65-IX). Cambridge, MA: Health
Effects Institute. http://pubs.healtheffects.org/view.php?id=82
Pinkerton. KE: Weller. BL; Menache. MG: Plopper. CG. (1998). Consequences of prolonged inhalation of
ozone on F344/N rats: Collaborative studies: Part XIII. A comparison of changes in the tracheobronchial
epithelium and pulmonary acinus in male rats at 3 and 20 months. (HEI Research Report 65). Cambridge,
MA: Health Effects Institute.
Plopper. CG: Chu. FP; Haselton. CJ: Peake. J: Wu. J: Pinkerton. KE. (1994). Dose-dependent tolerance to
ozone: I. Tracheobronchial epithelial reorganization in rats after 20 months' exposure. Am J Pathol 144:
404-420.
Plopper. CG: Smiley-Jewell. SM; Miller. LA; Fanucchi. MV; Evans. MJ; Buckpitt. AR; Avdalovic. M;
Gershwin. LJ; Joad. JP; Kaiekar. R; Larson. S: Pinkerton. KE: Van Winkle. LS: Schelegle. ES: Pieczarka.
EM; Wu. R; Hyde. DM. (2007). Asthma/allergic airways disease: Does postnatal exposure to
environmental toxicants promote airway pathobiology? Toxicol Pathol 35: 97-110.
http://dx.doi.org/10.1080/01926230601132030
Pope. CA. III. (2007). Mortality effects of longer term exposures to fine paniculate air pollution: Review of
recent epidemiological evidence [Review]. Inhal Toxicol 19: 33-38.
Pope. CA. Ill; Burnett. RT; Thun. MJ: Calle. EE; Krewski. D; Ito. K; Thurston. GD. (2002). Lung cancer,
cardiopulmonary mortality, and long-term exposure to fine paniculate air pollution. JAMA 287: 1132-1141.
Qian. Z; Liao. D: Lin. HM; Whitsel. EA; Rose. KM; Duan. Y. (2005). Lung function and long-term exposure
to air pollutants in middle-aged American adults. Arch Environ Occup Health 60: 156-163.
7-104
-------
Rage. E; Jacquemin. B; Nadif. R; Oryszczyn. MP; Siroux. V; Aguilera. I; Kauffmann. F; Kunzli. N. (2009a).
Total serum IgE levels are associated with ambient ozone concentration in asthmatic adults. Allergy 64: 40-
46. http://dx.doi.0rg/10.llll/i.1398-9995.2008.01800.x
Rage. E; Siroux. V; Kunzli. N; Pin. I; Kauffmann. F. (2009b). Air pollution and asthma severity in adults.
Occup Environ Med 66: 182-188. http://dx.doi.org/10.1136/oem.2007.038349
Rahman. I; Biswas. SK; Kode. A. (2006). Oxidant and antioxidant balance in the airways and airway diseases.
Eur J Pharmacol 533: 222-239. http://dx.doi.Org/10.1016/j.ejphar.2005.12.087
Ramadour. M; Burel. C: Lanteaume. A; Vervloet. D; Charpin. D; Brisse. F; Dutau. H; Charpin. D. (2000).
Prevalence of asthma and rhinitis in relation to long-term exposure to gaseous air pollutants. Allergy 55:
1163-1169.
Reiser. KM; Tyler. WS: Hennessv. SM; Dominguez. JJ; Last. JA. (1987). Long-term consequences of exposure
to ozone: II. Structural alterations in lung collagen of monkeys. Toxicol Appl Pharmacol 89: 314-322.
http://dx.doi.org/10.1016/0041-008X(87)90151-7
Renzetti. G: Silvestre. G: D'Amario. C: Bottini. E; Gloria-Bottini. F; Bottini. N; Auais. A; Perez. MK;
Piedimonte. G. (2009). Less air pollution leads to rapid reduction of airway inflammation and improved
airway function in asthmatic children. Pediatrics 123: 1051-1058. http://dx.doi.org/10.1542/peds.2008-1153
Ritz. B; Wilhelm. M. (2008). Ambient air pollution and adverse birth outcomes: Methodologic issues in an
emerging field. Basic Appl Ecol 102: 182-190.
Ritz. B; Wilhelm. M; Hoggatt. KJ; Ghosh. JK. (2007). Ambient air pollution and preterm birth in the
environment and pregnancy outcomes study at the University of California, Los Angeles. Am J Epidemiol
166: 1045-1052.
Ritz. B; Wilhelm. M; Zhao. Y. (2006). Air pollution and infant death in southern California, 1989-2000.
Pediatrics 118: 493-502.
Ritz. B; Yu. F. (1999). The effect of ambient carbon monoxide on low birth weight among children born in
southern California between 1989 and 1993. Environ Health Perspect 107: 17-25.
Ritz, B; Yu. F; Chapa. G: Fruin. S. (2000). Effect of air pollution on preterm birth among children born in
Southern California between 1989 and 1993. Epidemiology 11: 502-511.
Ritz. B; Yu. F; Fruin. S: Chapa. G: Shaw. GM; Harris. JA. (2002). Ambient air pollution and risk of birth
defects in Southern California. Am J Epidemiol 155: 17-25.
Rivas-Arancibia. S: Dorado-Martinez. C: Borgonio-Perez. G: Hiriart-Urdanivia. M; Verdugo-Diaz. L; Duran-
Vazquez. A; Colin-Baranque. L; Avila-Costa. MR. (2000). Effects of taurine on ozone-induced memory
deficits and lipid peroxidation levels in brains of young, mature, and old rats. Environ Res 82: 7-17.
http://dx.doi.org/10.1006/enrs.1999.3996
Rivas-Arancibia. S: Guevara-Guzman. R; Lopez-Vidal. Y; Rodriguez-Martinez. E; Gomes. MZ; Angoa-Perez.
M; Raisman-Vozari. R. (2010). Oxidative stress caused by ozone exposure induces loss of brain repair in
the hippocampus of adult rats. Toxicol Sci 113: 187-197. http://dx.doi.org/10.1093/toxsci/kfp252
Rivas-Manzano. P; Paz. C. (1999). Cerebellar morphological alterations in rats induced by prenatal ozone
exposure. Neurosci Lett 276: 37-40.
Rojas-Martinez. R; Perez-Padilla. R: Olaiz-Fernandez. G: Mendoza-Alvarado. L; Moreno-Macias. H; Fortoul.
T; Mcdonnell. W: Loomis. D: Romieu. I. (2007). Lung function growth in children with long-term
exposure to air pollutants in Mexico City. Am J Respir Crit Care Med 176: 377-384.
http://dx.doi.org/10.1164/rccm.200510-1678OC
Romero-Velazquez. RM; Alfaro-Rodriguez. A; Gonzalez-Pina. R: Gonzalez-Maciel. A. (2002). Effect of ozone
prenatal exposure on postnatal development of cerebellum. Proc West Pharmacol Soc 45: 65-67.
Romero. R; Espinoza. J: Kusanovic. JP; Gotsch. F; Hassan. S: Erez. O; Chaiworapongsa. T; Mazor. M. (2006).
The preterm parturition syndrome. BJOG 113: 17-42. http://dx.doi.Org/10.llll/i.1471-0528.2006.01120.x
7-105
-------
Romieu. I; Ramirez-Aguilar. M; Moreno-Macias. H; Barraza-Villarreal. A; Miller. P; Hernandez-Cadena. L;
Carbaial-Arroyo. LA; Hernandez-Avila. M. (2004a). Infant mortality and air pollution: Modifying effect by
social class. J Occup Environ Hyg 46: 1210-1216.
Rubes. J: Selevan. SG: Evenson. DP; Zudova. D; Vozdova. M; Zudova. Z; Robbins. WA; Perreault. SD.
(2005). Episodic air pollution is associated with increased DNA fragmentation in human sperm without
other changes in semen quality. Hum Reprod 20: 2776-2783.
Rubio. C: Paz. C. (2003). Indomethacin reverts sleep disorders produced by ozone exposure in rats. Toxicology
191: 89-96. http://dx.doi.org/10.1016/S0300-483X(03)00245-2
Salam. MT; Islam. T; Gauderman. WJ: Gilliland. FD. (2009). Roles of arginase variants, atopy, and ozone in
childhood asthma. J Allergy Clin Immunol 123: 596-602. http://dx.doi.Org/10.1016/i.jaci.2008.12.020
Salam. MT; Millstein. J: Li. YF; Lurmann. FW: Margolis. HG: Gilliland. FD. (2005). Birth outcomes and
prenatal exposure to ozone, carbon monoxide, and paniculate matter: Results from the Children's Health
Study. Environ Health Perspect 113: 1638-1644. http://dx.doi.org/10.1289/ehp.8111
Santiago-Lopez. D; Bautista-Martinez. JA; Reyes-Hernandez. CI; Aguilar-Martinez. M; Rivas-Arancibia. S.
(2010). Oxidative stress, progressive damage in the substantia nigra and plasma dopamine oxidation, in rats
chronically exposed to ozone. Toxicol Lett 197: 193-200. http://dx.doi.Org/10.1016/i.toxlet.2010.05.020
Santucci. D: Sorace. A; Francia. N; Aloe. L; Alleva. E. (2006). Prolonged prenatal exposure to low-level ozone
affects aggressive behaviour as well as NGF and BDNF levels in the central nervous system of CD-I mice.
Behav Brain Res 166: 124-130. http://dx.doi.Org/10.1016/i.bbr.2005.07.032
Schelegle. ES: Miller. LA: Gershwin. LJ: Fanucchi. MV: Van Winkle. LS: Gerriets. JE: Walbv. WF: Mitchell.
V: Tarkington. BK: Wong. VJ: Baker. GL: Pantle. LM: Joad. JP: Pinkerton. KE: Wu. R: Evans. MJ: Hyde.
DM; Plopper. CG. (2003). Repeated episodes of ozone inhalation amplifies the effects of allergen
sensitization and inhalation on airway immune and structural development in Rhesus monkeys. Toxicol
Appl Pharmacol 191: 74-85.
Schmelzer. KR; Wheelock. AM; Dettmer. K; Morin. D; Hammock. BD. (2006). The role of inflammatory
mediators in the synergistic toxicity of ozone and 1-nitronaphthalene in rat airways. Environ Health
Perspect 114: 1354-1360.
Schopke. R; Wolfer. DP; Lipp. HP; Leisinger-Trigona. MC. (1991). Swimming navigation and structural
variations of the infrapyramidal mossy fibers in the hippocampus of the mouse. Hippocampus 1: 315-328.
http://dx.doi.org/10.1002/hipo.450010322
Schroer. KT; Biagini Myers. JM; Ryan. PH; Lemasters. GK; Bernstein. PI; Villareal. M; Lockev. JE; Reponen.
T; Grinshpun. S: Khurana Hershev. GK. (2009). Associations between multiple environmental exposures
and Glutathione S-Transferase PI on persistent wheezing in a birth cohort. J Pediatr 154: 401-408,
408.e401. http://dx.doi.0rg/10.1016/i.ipeds.2008.08.040
Selevan. SG: Borkovec. L; Slott. VL; Zudova. Z; Rubes. J: Evenson. DP; Perreault. SD. (2000). Semen quality
and reproductive health of young Czech men exposed to seasonal air pollution. Environ Health Perspect
108: 887-894.
Sharkhuu. T; Doerfler. PL; Copeland. C: Luebke. RW: Gilmour. MI. (2011). Effect of maternal exposure to
ozone on reproductive outcome and immune, inflammatory, and allergic responses in the offspring. J
Immunotoxicol 8: 183-194. http://dx.doi.org/10.3109/1547691X.2011.568978
Silverman. RA; Ito. K. (2010). Age-related association of fine particles and ozone with severe acute asthma in
New York City. J Allergy Clin Immunol 125: 367-373. http://dx.doi.Org/10.1016/i.jaci.2009.10.061
Simonian. NA; Coyle. JT. (1996). Oxidative stress in neurodegenerative diseases [Review]. Annu Rev
Pharmacol Toxicol 36: 83-106. http://dx.doi.org/10.1146/annurev.pa.36.040196.000503
Slama. R: Darrow. L; Parker. J: Woodruff. TJ; Strickland. M; Nieuwenhuijsen. M; Glinianaia. S: Hoggatt. KJ;
Kannan. S: Hurley. F; Kalinka. J: Sram. R; Brauer. M; Wilhelm. M; Heinrich. J: Ritz. B. (2008). Meeting
report: Atmospheric pollution and human reproduction. Environ Health Perspect 116: 791-798.
7-106
-------
Smith. KR; Jerrett. M; Anderson. HR; Burnett. RT; Stone. V: Dement. R; Atkinson. RW: Cohen. A; Shonkoff.
SB; Krewski. D; Pope. CA. Ill; Thun. MJ; Thurston. G. (2009a). Public health benefits of strategies to
reduce greenhouse-gas emissions: Health implications of short-lived greenhouse pollutants. Lancet 374:
2091-2103. http://dx.doi.org/10.1016/s0140-6736(09)61716-5
Sokol. RZ; Kraft. P; Fowler. IM; Mamet. R; Kim. E; Berhane. KT. (2006). Exposure to environmental ozone
alters semen quality. Environ Health Perspect 114: 360-365. http://dx.doi.org/10.1289/ehp.8232
Son. JY; Cho. YS: Lee. JT. (2008). Effects of air pollution on postneonatal infant mortality among firstborn
infants in Seoul, Korea: Case-crossover and time-series analyses. Arch Environ Occup Health 63: 108-113.
http://dx.doi.Org/10.3200/AEOH.63.3.108-113
Sousa. SI; Ferraz. C: Alvim-Ferraz. MC: Martins. FG: Vaz. LG: Pereira. MC. (2011). Spirometric tests to
assess the prevalence of childhood asthma at Portuguese rural areas: Influence of exposure to high ozone
levels. Environ Int 37: 474-478. http://dx.doi.Org/10.1016/i.envint.2010.ll.014
Sousa. SIV: Alvim-Ferraz. MCM; Martins. FG: Pereira. MC. (2009). Ozone exposure and its influence on the
worsening of childhood asthma. Allergy 64: 1046-1055. http://dx.doi.org/10.1111/i. 1398-
9995.2009.01946.x
Sousa. SIV: Pereira. MMC: Martins. FG: Alvim-Ferraz. CM. (2008). Identification of regions with high ozone
concentrations aiming the impact assessment on childhood asthma. Hum Ecol Risk Assess 14: 610-622.
http://dx.doi.org/10.1080/10807030802074147
Sram. RJ; Binkova. B; Rossner. P; Rubes. J: Topinka. J: Dejmek. J. (1999). Adverse reproductive outcomes
from exposure to environmental mutagens. Mutat Res-Fundam Mol Mech Mutagen 428: 203-215.
http://dx.doi.org/10.1016/S1383-5742(99)00048-4
Stedman. JR; Kent. AJ. (2008). An analysis of the spatial patterns of human health related surface ozone
metrics across the UK in 1995, 2003 and 2005. Atmos Environ 42: 1702-1716.
http://dx.doi.0rg/10.1016/i.atmosenv.2007.ll.033
Stephens. RJ: Sloan. MF; Groth. DG. (1976). Effects of long-term, low-level exposure of NO2 or O3 on rat
lungs. Environ Health Perspect 16: 178-179.
Stockstill. BL; Chang. LY; Menache. MG: Mellick. PW: Mercer. RR; Crapo. JD. (1995). Bronchiolarized
metaplasia and interstitial fibrosis in rat lungs chronically exposed to high ambient levels of ozone. Toxicol
Appl Pharmacol 134: 251-263. http://dx.doi.org/10.1006/taap.1995.1191
Stokinger. HE. (1962). Effects of air pollution in animals. In AC Stern (Ed.), Air pollution (pp. 282-334). New
York, NY: Academic Press.
Strickland. MJ: Klein. M; Correa. A; Reller. MD; Mahle. WT; Riehle-Colarusso. TJ; Botto. LD; Flanders. WD;
Mulholland. JA; Siffel. C: Marcus. M; Tolbert. PE. (2009). Ambient air pollution and cardiovascular
malformations in Atlanta, Georgia, 1986-2003. Am J Epidemiol 169: 1004-1014.
http://dx.doi.org/10.1093/aie/kwp011
Tager. IB; Balmes. J: Lurmann. F; Ngo. L; Alcorn. S: Kunzli. N. (2005). Chronic exposure to ambient ozone
and lung function in young adults. Epidemiology 16: 751-759.
http://dx.doi.org/10.1097/01.ede.0000183166.68809.bO
Tamer. L; Calikoglu. M; Ates. NA; Yildirim. H; Ercan. B; Saritas. E; Unlu. A; Atik. U. (2004). Glutathione-S-
transferase gene polymorphisms (GSTT1, GSTM1, GSTP1) as increased risk factors for asthma.
Respirology 9: 493-498. http://dx.doi.Org/10.llll/j.1440-1843.2004.00657.x
Tovalin. H; Valverde. M; Morandi. MT; Blanco. S: Whitehead. L; Rojas. E. (2006). DNA damage in outdoor
workers occupationally exposed to environmental air pollutants. Occup Environ Med 63: 230-236.
Iran. MU; Weir. AJ: Fanucchi. MV; Rodriguez. AE; Pantle. LM; Smiley-Jewell. SM; Van Winkle. LS: Evans.
MJ: Miller. LA; Schelegle. ES: Gershwin. LJ; Hyde. DM; Plopper. CG. (2004). Smooth muscle
hypertrophy in distal airways of sensitized infant rhesus monkeys exposed to house dust mite allergen. Clin
Exp Allergy 34: 1627-1633. http://dx.doi.Org/10.llll/i.1365-2222.2004.02057.x
7-107
-------
Tsai. SS: Chen. CC: Hsieh. HJ; Chang. CC: Yang. CY. (2006). Air pollution and postneonatal mortality in a
tropical city: Kaohsiung, Taiwan. Inhal Toxicol 18: 185-189.
Tyler. WS: Tyler. NK; Last. JA; Gillespie. MJ; Barstow. TJ. (1988). Comparison of daily and seasonal
exposures of young monkeys to ozone. Toxicology 50: 131-144.
U.S. EPA (U.S. Environmental Protection Agency). (1996o). Table 6-19. Alveolar/bronchiolar tumor incidence
in B6C3F mice in the National Toxicology Program's chronic ozone study [EPA Report]. In Air quality
criteria for ozone and related photochemical oxidants (pp. 116). (EPA/600/P-93/004AF). Research Triangle
Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (2006b). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/R-05/004AF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
Van Bree. L; Dormans. JAM. A; Koren. HS: Devlin. RB; Rombout. PJA. (2002). Attenuation and recovery of
pulmonary injury in rats following short-term, repeated daily exposure to ozone. Inhal Toxicol 14: 883-900.
httrj://dx.doi.org/10.1080/08958370290084674
Van Winkle. LS: Baker. GL; Chan. JK; Schelegle. ES: Plopper. CG. (2010). Airway mast cells in a rhesus
model of childhood allergic airways disease. Toxicol Sci 116: 313-322.
http://dx.doi.org/10.1093/toxsci/kfqll9
Vanguilder. HP; Freeman. WM. (2011). The hippocampal neuroproteome with aging and cognitive decline:
Past progress and future directions. Front Aging Neurosci 3: 8. http://dx.doi.org/10.3389/fnagi.2011.00008
Veninga. TS. (1967). Toxicity of ozone in comparison with ionizing radiation. Strahlentherapie 134: 469-477.
von Mutius. E. (2009). Gene-environment interactions in asthma [Review]. J Allergy Clin Immunol 123: 3-11.
http://dx.doi.0rg/10.1016/i.jaci.2008.10.046
Vrijheid. M; Martinez. D; Manzanares. S: Dadvand. P; Schembari. A; Rankin. J; Nieuwenhuijsen. M. (2011).
Ambient air pollution and risk of congenital anomalies: A systematic review and meta-analysis [Review].
Environ Health Perspect 119: 598-606. http://dx.doi.org/10.1289/ehp.1002946
Wang. TN; Ko. YC: Qiao. YY; Huang. CC: Lin. RS. (1999). Association between indoor and outdoor air
pollution and adolescent asthma from 1995 to 1996 in Taiwan. Environ Res 81: 239-247.
Wang. XY; Hu. W: Tong. S. (2009c). Long-term exposure to gaseous air pollutants and cardio-respiratory
mortality in Brisbane, Australia. Geospat Health 3: 257-263.
Wenten. M; Gauderman. WJ: Berhane. K; Lin. PC: Peters. J: Gilliland. FD. (2009). Functional variants in the
catalase and myeloperoxidase genes, ambient air pollution, and respiratory-related school absences:
An example of epistasis in gene-environment interactions. Am J Epidemiol 170: 1494-1501.
http://dx.doi.org/10.1093/aie/kwp310
Wilhelm. M; Ritz. B. (2005). Local variations in CO and particulate air pollution and adverse birth outcomes
in Los Angeles County, California, USA. Environ Health Perspect 113: 1212-1221.
http://dx.doi.org/10.1289/ehp.7751
Witschi. H. (1991). Effects of oxygen and ozone on mouse lung tumorigenesis [Review]. Exp Lung Res 17:
473-483. http://dx.doi.org/10.3109/01902149109064433
Witschi. H; Espiritu. I; Pinkerton. KE; Murphy. K; Maronpot RR. (1999). Ozone carcinogenesis revisited.
Toxicol Sci 52: 162-167.
Witschi. H; Wilson. DW: Plopper. CG. (1993). Modulation of N-nitrosodiethylamine-induced hamster lung
tumors by ozone. Toxicology 77: 193-202.
Wollmann. HA. (1998). Intrauterine growth restriction: Definition and etiology. Horm Res 49: 1-6.
Wood. AM; Harrison. RM; Semple. S: Ayres. JG: Stocklev. RA. (2009). Outdoor air pollution is associated
with disease severity in al-antitrypsin deficiency. Eur Respir J 34: 346-353.
http://dx.doi.org/10.1183/09031936.00087908
7-108
-------
Woodruff. TJ; Darrow. LA; Parker. JD. (2008). Air pollution and postneonatal infant mortality in the United
States, 1999-2002. Environ Health Perspect 116: 110-115. http://dx.doi.org/10.1289/ehp.10370
Woodruff. TJ; Parker. JD; Adams. K; Bell. ML; Gehring. U; Glinianaia. S: Ha. EH; Jalaludin. B; Slama. R.
(2010). International Collaboration on Air Pollution and Pregnancy Outcomes (ICAPPO). Int J Environ Res
Public Health 7: 2638-2652. http://dx.doi.org/10.3390/iierph7062638
Woodruff. TJ: Parker. JD: Darrow. LA; Slama. R; Bell. ML; Choi. H; Glinianaia. S: Hoggatt. KJ; Karr. CJ:
Lobdell. DT; Wilhelm. M. (2009). Methodological issues in studies of air pollution and reproductive health.
Environ Res 109: 311-320. http://dx.doi.Org/10.1016/i.envres.2008.12.012
Yang. CY; Hsieh. HJ; Tsai. SS: Wu. TN; Chiu. HF. (2006). Correlation between air pollution and postneonatal
mortality in a subtropical city: Taipei, Taiwan. J Toxicol Environ Health A 69: 2033-2040.
http://dx.doi.org/10.1080/15287390600746181
Zanobetti. A; Bind. MAC: Schwartz. J. (2008). Paniculate air pollution and survival in a COPD cohort.
Environ Health Perspect 7: 48. http://dx.doi.org/10.1186/1476-069X-7-48
Zanobetti. A; Schwartz. J. (2007). Paniculate air pollution, progression, and survival after myocardial
infarction. Environ Health Perspect 115: 769-775. http://dx.doi.org/10.1289/ehp.9201
Zanobetti. A; Schwartz. J. (2011). Ozone and survival in four cohorts with potentially predisposing diseases.
Am J Respir Crit Care Med 184: 836-841. http://dx.doi.org/10.1164/rccm.201102-0227OC
Zelac. RE; Cromroy. HL; Bolch. WE. Jr; Dunavant. BG: Bevis. HA. (1971a). Inhaled ozone as a mutagen: I
Chromosome aberrations induced in Chinese hamster lymphocytes. Environ Res 4: 262-282.
Zelac. RE; Cromroy. HL; Bolch. WE. Jr; Dunavant. BG: Bevis. HA. (1971b). Inhaled ozone as a mutagen: II
Effect on the frequency of chromosome aberrations observed in irradiated Chinese hamsters. Environ Res
4: 325-342.
7-109
-------
8 POPULATIONS POTENTIALLY AT INCREASED RISK FOR
OZONE-RELATED HEALTH EFFECTS
Interindividual variation in human responses to air pollution exposure can result in
some groups being at increased risk for detrimental effects in response to ambient
exposure to an air pollutant. The NAAQS are intended to provide an adequate margin
of safety for both the population as a whole and those potentially at increased risk for
health effects in response to ambient air pollution1. To facilitate the identification of
populations and lifestages at greater risk for air pollutant related health effects,
studies have evaluated factors that may contribute to the susceptibility and/or
vulnerability of an individual to air pollutants. The definitions of susceptibility and
vulnerability have been found to vary across studies, but in most instances
"susceptibility" refers to biological or intrinsic factors (e.g., lifestage, sex, pre-
existing disease/conditions) while "vulnerability" refers to non-biological or extrinsic
factors (e.g., socioeconomic status [SES]) (U.S. EPA. 2010c. 2009d). In some cases,
the terms "at-risk" and "sensitive" populations have been used to encompass these
concepts more generally. The main goal of this evaluation is to identify and
understand those factors that result in a population or lifestage being at increased risk
of an air pollutant-related health effect, not to categorize the factors. To this end,
previous ISAs and reviews (Sacks et al.. 2011: U.S. EPA. 2010c. 2009d) have used
"susceptible populations" to encompass these various factors. In this chapter,
"at-risk" is the all-encompassing term used for groups with specific factors that
increase the risk of an air pollutant (e.g., O3)-related health effect in a population.
Individuals, and ultimately populations, could experience increased risk for air
pollutant induced health effects in multiple different ways. A group with intrinsically
increased risk would have some factor(s) that increases risk for an effect through
a biological mechanism. In general, people in this category would have a steeper
concentration-risk relationship, compared to those not in the category. Potential
factors that are often considered intrinsic include genetic background and sex.
A group of people could also have extrinsically increased risk, which would be
through an external, non-biological factor. Examples of extrinsic factors include
SES and diet.
In addition, some groups are at increased risk due to differential exposure, which can
encompass multiple forms. This includes increased risk due to increased internal
dose at a given exposure concentration. For example, individuals may have a greater
dose of delivered pollutant because of their breathing pattern. This group would
include persons who work outdoors or exercise outdoors. Some outdoor workers
could also have greater exposure (concentration x time), regardless of the delivered
dose and this greater exposure may increase the risk of O3-related health effects.
1 The legislative history of section 109 indicates that a primary standard is to be set at "the maximum permissible ambient air
level ... which will protect the health of any [sensitive] group of the population," and that for this purpose "... reference should be
made to a representative sample of persons comprising the sensitive group rather than to a single person in such a group."
[S. Rep. No. 91-1196, 91st Cong., 2d Sess. 10(1970)].
8-1
-------
Finally, there are those who might be placed at increased risk for experiencing a
greater exposure, and therefore increased risk of health effects, by being exposed at a
higher concentration. For example, groups of people exposed to higher air pollutant
concentrations due to less availability/use of home air conditioners (i.e., more open
windows on high O3 days) or close proximity to known sources of air pollution.
Some factors described above are multifaceted. For example, SES may affect access
to medical care, which itself may contribute to the presence of pre-existing diseases
and conditions considered as intrinsic factors. Additionally, children tend to spend
more time outdoors at higher levels of activity than adults, which leads to increased
intake dose and exposure, but they also have biological (i.e., intrinsic) differences
when compared to adults.
The emphasis of this chapter is to identify and understand the factors that potentially
increase the risk of O3-related health effects, regardless of whether the increased risk
is due to intrinsic factors, extrinsic factors, increased dose, increased exposure, or a
combination. The following sections examine factors that potentially lead to
increased risk of O3-related health effects and characterize the overall weight of
evidence for each factor. Most of the factors are related to greater health effects given
a specific dose but there is also discussion of increased internal dose and/or exposure
at a given concentration integrated throughout the sections (i.e., lifestage, outdoor
workers, and air conditioning use).
Approach to Classifying Potential At-Risk Factors
To identify factors that potentially lead to some populations being at greater risk to
air pollutant related health effects, the evidence across relevant scientific disciplines
(i.e., exposure sciences, dosimetry, controlled human exposure, toxicology, and
epidemiology) was evaluated. In this systematic approach, the collective evidence is
used to examine coherence of effects across disciplines and determine biological
plausibility. By first focusing on studies that conduct stratified analyses
(i.e., epidemiologic or controlled human exposure) it is possible to identify factors
that may result in some populations being at greater risk of an air pollutant related
health effect. These types of studies allow for an evaluation of populations exposed
to similar air pollutant (e.g., O3) concentrations within the same study design.
Experimental studies also provide important lines of evidence in the evaluation of
factors that may lead to increased risk of an air pollutant related-health effect.
Toxicological studies conducted using animal models of disease and controlled
human exposure studies that examine individuals with underlying disease or genetic
polymorphisms may provide evidence to inform whether a population is at increased
risk of an air pollutant related health effect in the absence of stratified epidemiologic
analyses. Additionally these studies can provide support for coherence with the
health effects observed in epidemiologic studies as well as an understanding of
biological plausibility. Information on factors that may result in increased risk of O3-
related health effects can also be obtained from studies that examine exposure
differences between populations. The collective results across the scientific
8-2
-------
disciplines comprise the overall weight of evidence that is used to determine whether
a specific factor results in a population being at increased risk of an air pollutant
related health effect.
Building on the causal framework discussed in detail in the Preamble and used
throughout the ISA, conclusions are made regarding the strength of evidence, based
on the evaluation and synthesis across scientific disciplines, for each factor that may
contribute to increased risk of an O3-related health effect. The conclusions were
drawn while considering the "Aspects to Aid in Judging Causality" discussed in
Table 1 of the Preamble. The categories considered for evaluating the potential
increased risk of an air pollutant-related health effect are "adequate evidence,"
"suggestive evidence," "inadequate evidence," and "evidence of no effect." They are
described in more detail in Table 8-1.
Table 8-1
Classification of Evidence for Potential At-Risk Factors.
Suggestive
evidence
Health Effects
There is substantial, consistent evidence within a discipline to conclude that a factor results in a population or
Adequate lifestage being at increased risk of air pollutant-related health effect(s) relative to some reference population or
evidence lifestage. Where applicable this includes coherence across disciplines. Evidence includes multiple high-quality
studies.
The collective evidence suggests that a factor results in a population or lifestage being at increased risk of an air
pollutant-related health effect relative to some reference population or lifestage, but the evidence is limited due to
some inconsistency within a discipline or, where applicable, a lack of coherence across disciplines.
Inadequate
evidence
The collective evidence is inadequate to determine if a factor results in a population or lifestage being at increased
risk of an air pollutant-related health effect relative to some reference population or lifestage. The available studies
are of insufficient quantity, quality, consistency and/or statistical power to permit a conclusion to be drawn.
There is substantial, consistent evidence within a discipline to conclude that a factor does not result in a population
Evidence of or lifestage being at increased risk of air pollutant-related health effect(s) relative to some reference population or
no effect lifestage. Where applicable this includes coherence across disciplines. Evidence includes multiple high-quality
studies.
This chapter evaluates the various factors indicated in the literature that may result in
a population being at increased risk of an O3-related health effect. For further detail
on the epidemiologic, controlled human exposure, and toxicological studies included
in this chapter, see Chapters 5., 6, and 7.
8.1 Genetic Factors
The potential effects of air pollution on individuals with specific genetic
characteristics have been examined; studies often target polymorphisms in already
identified candidate susceptibility genes or in genes whose protein products are
thought to be involved in the biological mechanism underlying the health effect of an
air pollutant (Sacks et al. 2011). As a result, multiple studies that examined the
effect of short- and long-term O3 exposure on respiratory function have focused on
whether various gene profiles lead to an increased risk of O3-related health effects.
8-3
-------
For more details on the function and mode of action of the genetic factors discussed
in this section, see Section 5.4.2.1. Additionally, a limited number of toxicological
studies have examined the joint effects of nutrition and genetics. Details on these
toxicological studies of nutrition and genetics can be found in Section 5.4.2.3.
Multiple genes, including glutathione S-transferase Mu 1 (GSTM1) and tumor
necrosis factor-a (TNF-a) were evaluated in the 2006 O3 AQCD and found to have a
"potential role... in the innate susceptibility to O3" (U.S. EPA, 2006b).
Epidemiologic, controlled human exposure, and toxicological studies performed
since the 2006 O3 AQCD have continued to examine the roles of GSTM1 and TNF-a
in modifying O3-related health effects and have examined other gene variants that
may also increase risk. Due to small sample sizes, many controlled human exposure
studies are limited in their ability to test genes with low frequency minor alleles and
therefore, some genes important for O3-related health effects may not have been
examined in these types of studies. A summary of effect measure modification
findings from epidemiologic and controlled human exposure studies discussed in this
section is included in Table 8-2 and from animal toxicology studies in Table 8-3.
Epidemiologic studies that examined the effects of short-term exposure to O3 on lung
function included analyses of potential gene-environment interactions. Romieu et al.
(2006) reported an association between O3 and respiratory symptoms that were larger
among children with GSTM1 null or glutathione S-transferase P 1 (GSTP1) Val/Val
genotypes compared with children with GSTM1 positive or GSTP1 lie/lie or Ile/Val
genotypes, respectively. However, results suggested that O3-associated decreases in
lung function may be greater among children with GSTP1 He/Tie or Ile/Val compared
to GSTP1 Val/Val. Alexeeff et al. (2008) reported greater O3-related decreases in
lung function among GSTP1 Val/Val adults than those with GSTP1 He/Tie or GSTP1
Ile/Val genotypes. In addition, they detected greater O3-associated decreases in lung
function for adults with long GT dinucleotide repeats in heme-oxygenase-1
(HMOX1) promoters.
8-4
-------
Table 8-2 Summaries of results from epidemiologic and controlled human
exposures studies of modification by genetic variants.
Gene variant
GSTM1 null
GSTP1 Val/Val
GSTP1 lie/lie
or Ile/Val
GSTP1 Ile/Val
or Val/Val
HMOX1 S/L or L/L
NQO1 wildtype
and GSTM1 null
NQO1 wildtype
and GSTM1 null
NQO1 wildtype
and GSTM1 null
Comparison group
GSTM1 positive
GSTP1 lie/lie or Ile/Val
GSTP1 Val/Val
GSTP1 lie/lie
HMOX1 S/S
Other combinations
Other combinations
Other combinations
Health outcome
/population
Respiratory symptoms among
asthmatic children
Respiratory symptoms among
asthmatic children
Lung function among
asthmatic children
Lung function among adults
Lung function among adults
Lung function among healthy
adults with exercise
Lung function among mild-to-
moderate asthmatics with
moderate exercise
Inflammatory responses
among mild-to-moderate
asthmatics with moderate
exercise
Effect modification of
association for the
gene variant
t
t
4
4
4
4
=
Reference
Romieu et al. (2006)
Alexeeffetal. (2008)
Bergamaschi et al.
(2001 )
Vaaaggini etal. (2010)
GSTM1 null
GSTM1 positive
Lung function among healthy
adults with intermittent
moderate exercise
GSTM1 null
GSTM1 positive
Inflammatory changes among
healthy adults with intermittent
moderate exercise
Kim etal. (2011)
GSTM1 null
GSTM1 null
GSTM1 null
GSTM1 positive
GSTM1 positive
GSTM1 positive
Inflammatory responses
among healthy adults with
intermittent moderate exercise
Lung function among
asthmatic children
Lung function among healthy
adults with intermittent
moderate exercise
-
J, Romieu et al. (2004b)
Alexis etal. (2009)
8-5
-------
Several controlled human exposure studies have reported that genetic polymorphisms
of antioxidant enzymes may modulate pulmonary function and inflammatory
responses to O3 challenge. Healthy carriers of NAD(P)H quinone oxidoreductase 1
(NQO1) wild type (wt) in combination with GSTM1 null genotype had greater
decreases in lung function parameters with exposure to O3 (Bergamaschi et al.,
2001). Vagaggini et al. (2010) exposed mild-to-moderate asthmatics to O3 during
moderate exercise. In subjects with NQO1 wt and GSTM1 null, there was no
evidence of changes in lung function or inflammatory responses to O3. Kim et al.
(2011) also recently conducted a study among young adults, about half of whom
were GSTMl-null and half of whom were GSTM1-sufficient. They detected no
difference in the FEVi responses to O3 exposure by GSTM1 genotype and did not
examine NQO1. In another study that examined GSTM1 but not NQO1, asthmatic
children with GSTM1 null genotype (Romieu et al.. 2004b) were reported to have
greater decreases in lung function in relation to O3 exposure. Additionally,
supplementation with antioxidants (Vitamins C and E) had a slightly more beneficial
effect among GSTM1 null children (for more on modification by diet, see
Section 8.4.1).
In a study of healthy volunteers with GSTM1 sufficient and GSTM1 null genotypes
exposed to O3 with exercise, Alexis et al. (2009) found genotype effects on
inflammatory responses but not lung function responses to O3. At 4 hours post-O3
exposure, individuals with either GSTM1 genotype had statistically significant
increases in sputum neutrophils with a tendency for a greater increase in GSTM1
sufficient than GSTM1 nulls. At 24 hours postexposure, neutrophils had returned to
baseline levels in the GSTM1 sufficient individuals. In the GSTM1 null subjects,
neutrophil levels increased from 4 to 24 hours and were significantly greater than
both baseline levels and levels at 24 hours in the GSTM1 sufficient individuals.
In addition, O3 exposure increased the expression of the surface marker CD 14 in
airway neutrophils of GSTM1 null subjects compared with GSTM1 sufficient
subjects. CD 14 and TLR4 are co-receptors for endotoxin, and signaling through this
innate immune pathway has been shown to be important for a number of biological
responses to O3 exposure in toxicological studies (Garantziotis et al.. 2010:
Hollingsworth et al.. 2010: Hollingsworth et al.. 2004: Kleeberger et al.. 2000).
Alexis et al. (2009) also demonstrated decreased numbers of airway macrophages at
4 and 24 hours following O3 exposure in GSTM1 sufficient subjects. Airway
macrophages in GSTM1 null subjects were greater in number and found to have
greater oxidative burst and phagocytic capability following O3 exposure than those
of GSTM1 sufficient subjects. Airway macrophages and dendritic cells from GSTM1
null subjects exposed to O3 expressed higher levels of the surface marker HLA-DR;
again suggesting activation of the innate immune system. Since there was no FA
control in the Alexis et al. (2009) study, effects of the exposure other than O3 cannot
be ruled out. In general, the findings between these studies are inconsistent. It is
possible that different genes may be important for different phenotypes. Additional
studies, which include appropriate controls, are needed to clarify the influence of
genetic polymorphisms on O3 responsiveness in humans.
8-6
-------
Table 8-3 Summaries of results from animal toxicology studies of
modification by genetic variants.
Gene
variant
Tlr4
Tlr2
MyD88
Tnfr1/Tnfr2
Nfkb
Jnk
II6
1110
Marco
Nos2
Hsp70
NQO1
Csb
Mmp9
CD44
Cxcr2
1113
Reference3
Hollingsworth et al. (2004):
Kleebergeretal. (2000)
Williams et al. (2007b)
Williams et al. (2007b)
Williams et al. (2007b)
Choetal.(2001)
Cho et al. (2007)
Cho et al. (2007)
Cho et al. (2007)
Johnston et al. (2005b)
Backus etal. (2010)
Dahl et al. (2007)
Kleebergeretal. (2001)
Fakhrzadeh et al. (2002)
Kenvon et al. (2002)
Bauer etal. (2011)
Vovnow et al. (2009)
Kooteretal. (2007)
Yoon et al. (2007)
Garantziotis et al. (2009)
Johnston et al. (2005)
Williams et al. (2008b)
Exposure
0.3 ppm, 72 hours
0.3 ppm, 72 hours
2.0 ppm, 3 hours
0.3 ppm, 3-24 hours
0.3 ppm, 3-24 hours
0.3 ppm, 3-24 hours
0.3 ppm, 3-48 hours
2.0 ppm, 3 hours
0.3 ppm, 6-48 hours
0.3 ppm, 6-48 hours
0.3 ppm, 3-72 hours
2.0 ppm, 3 hours
0.3 ppm, 24-
72 hours
0.3 ppm, 48 hours
0.3 ppm, 72 hours
0.8 ppm, 3 hours
1.0 ppm,
8 h/night for 3 nights
0.3 ppm, 6-72 hours
1 .0 ppm, 3 hours
0.8 ppm, 8 hours
0.3 ppm 6-72 hours
2.0 ppm, 3 hours
1.0 ppm, 3 hours
3.0 ppm, 3 hours
Health outcome /population
Decreased hyperpermeability
No genotype difference in hyperpermeability, BALF cells,
orAHRatO.Sppm. Reduced AHR at 2.0ppm.
Decreased AHR. Reduced inflammation at 3 hours
Decreased inflammation and AHR
Decreased inflammation, hyperpermeability, and AHR
Decreased BALF cells, neutrophilia and lung damage.
No genotype difference in hyperpermeability.
Reduced AHR
Decreased inflammation, hyperpermeability, and lung
damage
Decreased inflammation, hyperpermeability, and lung
damage
Decreased neutrophilia and hyperpermeability, reduced
soluble TNFRs, no effect on AHR
Reduced neutrophilia and soluble TNFR2 and MIP-2
Increased inflammation
Increased inflammation, 8-isoprostane, and
hyperpermeability
Decreased hyperpermeability, no effect on BALF cells
Reduced AM nitric oxide, reactive nitrogen species, and
superoxide anion, decreased PGE2, increased COX
expression, decreased hyperpermeability and BALF cells
Increased hyperpermeability, neutrophilia, MMP-9 activity,
and protein nitration products
Decreased hyperpermeability and inflammation
Reduced inflammation and AHR
Decreased TNF-a in BALF. No genotype difference in
neutrophilia or lung damage.
Increased hyperpermeability, neutrophilia, inflammation,
and lung damage
Decreased AHR
Reduced neutrophilia, lung injury, and AHR. No change in
chemokine expression or hyperpermeability
Reduced AHR, BALF cells, and neutrophilia
"The table includes animal toxicology studies where responses are assessed after gene deletion.
3-7
-------
In general, toxicological studies have reported differences in cardiac and respiratory
effects after O3 exposure among different mouse strains, which alludes to differential
risk among individuals due to genetic variability (Tankersley et al, 2010; Chuang et
al., 2009; Hamade and Tankersley, 2009; Hamade et al., 2008). Thus strains of mice
which are prone to or resistant to O3-induced effects have been used to
systematically identify candidate genes that may increase risk of O3-related health
effects. Genome wide linkage analyses have identified quantitative trait loci for
O3-induced lung inflammation and hyperpermeability on chromosome 17
(Kleeberger et al.. 1997) and chromosome 4 (Kleeberger et al.. 2000). respectively,
using recombinant inbred strains of mice. More specifically, these studies found that
TNF (protein product is the inflammatory cytokine TNF-a) and Tlr4 (protein product
is TLR4, involved in endotoxin responses) were candidate susceptibility genes
(Kleeberger et al.. 2000; Kleeberger et al.. 1997). The TNF receptors 1 and 2 have
also been found to play a role in injury, inflammation, and airway hyperreactivity in
studies of O3-exposed knockout mice (Cho et al.. 2007; Cho et al.. 2001) through
NF-KB and MAPK/AP-1 (Jnk) signaling pathways (Cho et al.. 2007). In addition to
Tlr4, other innate immune pattern recognition signaling pathway genes, including
Tlr2 and Myd88, appear to be important in responses to O3, as demonstrated by
Williams et al. (2007b). A role for the inflammatory cytokine IL-6 has been
demonstrated in gene-deficient mice with respect to inflammation and injury, but not
AHR (Johnston et al.. 2005b; Yu et al.. 2002). Other studies have demonstrated a key
role for CXCR2, the chemokine receptor for the neutrophil chemokines KC and
MIP-2, (Johnston et al.. 2005a) and CD44, the major receptor for the extracellular
matrix component hyaluronan (Garantziotis et al.. 2009) in O3-mediated AHR. Mice
deficient in IL-10, an anti-inflammatory cytokine, demonstrated increased pulmonary
inflammation in response to O3 exposure (Backus et al., 2010). Thus genes related to
innate immune signaling and pro- and anti-inflammatory genes are important for
O3-induced responses.
Altered O3 responses between mouse strains could be due to genetic variability in
nuclear factor erythroid 2-related factor 2 (Nrf-2), suggesting a role for genetic
differences in altering the formation of ROS (Hamade et al.. 2010). Additionally,
some studies have reported O3-related effects to vary by Inf-1 and Inf-2 quantitative
trait loci (Tankerslev and Kleeberger. 1994) and a gene coding for Clara cell
secretory protein (CCSP) (Broeckaert et al.. 2003; Wattiez et al.. 2003). Other
investigations in inbred mouse strains found that differences in expression of certain
proteins, such as CCSP (Broeckaert et al.. 2003) and MARCO (Pahl et al.. 2007). are
responsible for phenotypic characteristics, such as epithelial permeability and
scavenging of oxidized lipids, respectively, which confer sensitivity to O3.
Nitric oxide (NO), derived from activated macrophages, is produced upon exposure
to O3 and is thought to participate in lung damage. Mice deficient in the gene for
inducible nitric oxide synthase (NOS2/NOSII/iNOS) are partially protected against
lung injury (Kleeberger et al.. 2001). and it appears that O3-induced iNOS expression
is tied to the TLR4 pathway described above. Similarly, iNOS deficient mice do not
produce reactive nitrogen intermediates after O3 exposure, in contrast to their
wild-type counterparts, and also produce less PGE2 comparatively (Fakhrzadeh et
-------
al., 2002). These gene-deficient mice were protected from O3-induced lung injury
and inflammation. In contrast, another study using a similar exposure concentration
but longer duration of exposure found that iNOS deficient mice were more at risk of
Os-induced lung damage (Kenyon et al., 2002). Therefore, the role of iNOS in
mediating the response to O3 exposure is likely dependent on the exposure
concentration and duration.
Voynow et al. (2009) have shown that NQO1 deficient mice, like their human
counterparts, are resistant to O3-induced AHR and inflammation. NQO1 catalyzes
the reduction of quinones to hydroquinones, and is capable of both protective
detoxification reactions and redox cycling reactions resulting in the generation of
reactive oxygen species. Reduced production of inflammatory mediators and cells
and blunted AHR were observed in NQO1 null mice after exposure to O3. These
results correlated with those from in vitro experiments in which human bronchial
epithelial cells treated with an NQO1 inhibitor exhibited reduced inflammatory
responses to exposure to O3. This study may provide biological plausibility for the
increased biomarkers of oxidative stress and increased pulmonary function
decrements observed in O3-exposed individuals bearing both the wild-type NQO1
gene and the null GSTM1 gene (Bergamaschi et al., 2001). Deletion of the gene for
MMP9 also conferred protection against O3-induced airways inflammation and
injury (Yoon et al., 2007).
The role of TNF-a signaling in O3-induced responses has been previously
established through depletion experiments, but a more recent toxicological study
investigated the effects of combined O3 and PM exposure in transgenic TNF
overexpressing mice. Kumarathasan et al. (2005) found that subtle effects of these
pollutants were difficult to identify in the midst of the severe pathological changes
caused by constitutive TNF-a overexpression. However, there was evidence that
TNF transgenic mice were at increased risk of O3/PM-induced oxidative stress, and
they exhibited elevation of a serum creatine kinase after pollutant exposure, which
may suggest potential systemic or cardiac related effects. Differential risk of O3
among inbred strains of animals does not seem to be dose dependent since absorption
of 18O in various strains of mice did not correlate with resistance or sensitivity
(Vanczaetal..20Q9).
Defects in DNA repair mechanisms may also confer increased risk of O3-related
health effects. Cockayne syndrome, a rare autosomal recessive disorder in humans, is
characterized by UV sensitivity abnormalities, neurological abnormalities, and
premature aging. The same genetic defect in mice (Csb~'~) makes them sensitive to
oxidative stressors, including O3. Kooter et al. (2007) demonstrated that Csb~'~ mice
produced significantly more TNF-a after exposure to O3 than their wild-type
counterparts. However, there were no statistically significant differences in other
markers of inflammation or lung injury between the two strains of mice.
Overall, for variants in multiple genes there is adequate evidence for involvement in
populations being more at-risk than others to the effects of O3 exposure on health.
Controlled human exposure and epidemiologic studies have reported evidence of O3-
related increases in respiratory symptoms or decreases in lung function with variants
8-9
-------
including GSTM1, GSTP1, HMOX1, and NQO1. NQO1 deficient mice were found
to be resistant to O3-induced AHR and inflammation, providing biological
plausibility for results of studies in humans. Additionally, studies of rodents have
identified a number of other genes that may affect O3-related health outcomes,
including genes related to innate immune signaling and pro- and anti-inflammatory
genes, which have not been investigated in human studies.
8.2 Pre-existing Disease/Conditions
Individuals with certain pre-existing diseases are likely to constitute an at-risk
population. This may be the result of individuals with a pre-existing
disease/condition having less reserve than healthy individuals, so although the
absolute change may be the same, the health consequences are different. Previous O3
AQCDs concluded that some people with pre-existing pulmonary disease, especially
asthma, are among those at increased risk of an O3-related health effect. Extensive
toxicological evidence indicates that altered physiological, morphological and
biochemical states typical of respiratory diseases may render people at risk of an
additional oxidative burden induced by O3 exposure. In addition, a number of
epidemiologic studies found that some individuals with respiratory diseases are at
increased risk of O3-related effects. The majority of the studies identified in previous
AQCDs focused on whether pre-existing respiratory diseases result in increased risk
of O3-related health effects, with a limited number of studies examining other
pre-existing diseases, such as cardiovascular.
Studies identified since the completion of the 2006 O3 AQCD that examined whether
pre-existing diseases and conditions lead to increased risk of O3-induced health
effects were identified and are summarized below. Table 8-4 displays the prevalence
rates of some of these conditions categorized by age and region among adults in the
U.S. population; data for children, when available, are presented within the following
sections. Substantial proportions of the U.S. population are affected by these
conditions and therefore may represent potentially large at-risk populations. While
these diseases and conditions represent biological or intrinsic factors that could lead
to increased risk, the pathways to their development may have intrinsic or extrinsic
origins.
8-10
-------
Table 8-4 Prevalence of respiratory diseases, cardiovascular diseases, and
diabetes among adults by age and region in the U.S.
Adults
Chronic
Disease/Condition
N A e
(in thousands)
18-44 45-64 65-74 75+ Nort.h
63ST
Region
Mid
west
South West
Respiratory Diseases
Asthma3
16,380 7.2 7.5 7.8 6.4 7.7
8.0
5.9 8.4
COPD
Chronic Bronchitis
Emphysema
9,832 3.2 5.5 5.9 5.3 3.4
3,789 0.2 2.0 5.7 5.0 1.2
4.8
1.9
5.2 2.9
1.9 1.3
Cardiovascular
Diseases
All Heart Disease
Coronary Heart
Disease
Hypertension
Diabetes
26,628 4.6 12.3 26.7 39.2 11.3
14,428 1.1 6.7 16.9 26.7 5.7
56,159 8.7 32.5 54.4 61.1 22.9
18,651 2.3 12.1 20.4 17.3 4.5
12.7
6.5
24.1
7.6
12.2 9.9
7.3 4.9
27.1 20.6
9.0 7.7
aAsthma prevalence is reported for "still has asthma."
Source: Pleiset al. (2009): National Center for Health Statistics.
8.2.1 Influenza/Infections
Recent studies have indicated that underlying infections may increase the risk of O3-
related health effects because O3 exposure likely impairs host defenses, which may
increase the body's response to an infectious agent. However, there is little
epidemiologic or experimental evidence that infection or influenza itself renders an
individual at greater risk of an O3-induced health effect. A study of hospitalizations
in Hong Kong reported that increased levels of influenza intensity resulted in
increased excess risk of respiratory disease hospitalizations related to O3 exposure
(Wong et al.. 2009). In addition, a study of lung function in asthmatic children
reported decreases in lung function with increased short-term O3 exposure for those
with upper respiratory infections but not for those without infections (Lewis et al..
2005). Toxicological studies provide biological plausibility for the increase in
O3-induced health effects observed in epidemiologic studies that examined infections
by way of studies that demonstrated that exposure to 0.08 ppm O3 increased
streptococcus-induced mortality, regardless of whether O3 exposure preceded or
followed infection (Miller et al.. 1978: Coffin and Gardner. 1972: Coffin et al..
1967). Overall, the epidemiologic and experimental evidence supports the potential
for increased risk to be conferred by an infection but the number of studies is limited.
There have only been a few epidemiologic studies and these studies examine
8-11
-------
different outcomes (respiratory-related hospital admissions or lung function) and
different modifiers (influenza or respiratory infection). In some of the toxicological
studies, the O3 exposure came before the infection. Therefore, evidence is inadequate
to determine if influenza/infections increase the risk of O3-related health effects.
8.2.2 Asthma
Previous O3 AQCDs identified individuals with asthma as a population at increased
risk of O3-related health effects. Within the U.S., approximately 7.3% of adults have
reported currently having asthma (Pleis et al., 2009), and 9.5% of children have
reported currently having asthma (Bloom et al., 2008). For more detailed prevalence
by age, see Table 8-5.
Table 8-5 Prevalence of asthma by age in the U.S.
Age (years)
N (in thousands)
Percent
0-4
1,276
6.2
5-11
3,159
11.2
12-17
2,518
10.2
18-44
7,949
7.2
45-64
5,768
7.5
65-74
1,548
7.8
75+
1,116
6.4
aAsthma prevalence is reported for "still has asthma."
Source: Statistics for adults: Pleis et al. (2009): statistics for children: Bloom et al. (2008): National Center for Health Statistics.
Multiple epidemiologic studies included within this ISA have evaluated the potential
for increased risk of O3-related health effects among individuals with asthma.
A study of lifeguards in Texas reported decreased lung function with short-term O3
exposure among both individuals with and without asthma, however, the decrease
was greater among those with asthma (Thaller et al., 2008). A Mexican study of
children ages 6-14 detected an association between short-term O3 exposure and
wheeze, cough, and bronchodilator use among asthmatics but not non-asthmatics,
although this may have been the result of a small non-asthmatic population
(Escamilla-Nufiez et al., 2008). A study of modification by airway
hyperresponsiveness (AHR) (a condition common among asthmatics) reported
greater short-term O3-associated decreases in lung function in elderly individuals
with AHR, especially among those who were obese (Alexeeff et al., 2007). However,
no evidence for increased risk was found in a study performed among children in
Mexico City that examined the effect of short-term O3 exposure on respiratory health
(Barraza-Villarreal et al., 2008). In this study, a positive association was reported for
airway inflammation among asthmatic children, but the observed association was
8-12
-------
similar in magnitude to that of non-asthmatics. Similarly, a study of children in
California reported an association between O3 concentration and exhaled nitric oxide
fraction (FeNO) that persisted both among children with and without asthma as well
as those with and without respiratory allergy (Berhane et al., 2011). Finally, Khatri et
al. (2009) found no association between short-term O3 exposure and altered lung
function for either asthmatic or non-asthmatic adults, but did note a decrease in lung
function among individuals with allergies.
Evidence for difference in effects among asthmatics has been observed in studies that
examined the association between O3 exposure and altered lung function by asthma
medication use. A study of children with asthma living in Detroit reported a greater
association between short-term O3 and lung function for corticosteroid users
compared with noncorticosteroid users (Lewis et al., 2005). Conversely, another
study of children found decreased lung function among noncorticosteroid users
compared to corticosteroid users, although in this study, a large proportion of
non-users were considered to be persistent asthmatics (Hernandez-Cadena et al.,
2009). Lung function was not related to short-term O3 exposure among corticosteroid
users and non-users in a study taking place among children during the winter months
in Canada (Liu et al., 2009a). Additionally, a study of airway inflammation among
individuals aged 12-65 years old reported a counterintuitive inverse association with
O3 of similar magnitude for all groups of corticosteroid users and non-users (Qian et
al.. 2009).
Controlled human exposure studies that have examined the effects of O3 on
individuals with asthma and healthy controls are limited. Based on studies reviewed
in the 1996 and 2006 O3 AQCDs, subjects with asthma appeared to be at least as
sensitive to acute effects of O3 in terms of FEVi and inflammatory responses as
healthy non-asthmatic subjects. For instance, Horstman et al. (1995) observed that
mild-to-moderate asthmatics, on average, experienced double the O3-induced FEVi
decrement of healthy subjects (19% versus 10%, respectively, p = 0.04). Moreover, a
statistically significant positive correlation between FEVi responses to O3 exposure
and baseline lung function was observed in individuals with asthma, i.e., responses
increased with severity of disease. Kreit et al. (1989) performed a short duration
study in which asthmatics also showed a considerably larger average O3-induced
FEVi decrement than the healthy controls (25% versus 16%, respectively) following
exposure to O3 with moderate-heavy exercise. Alexis et al. (2000) and Jorres et al.
(1996) also reported a tendency for slightly greater FEVi decrements in asthmatics
than healthy subjects. Minimal evidence exists suggesting that individuals with
asthma have smaller O3-induced FEVi decrements than healthy subjects (3% versus
8%, respectively) (Mudway et al., 2001). However, the asthmatics in that study also
tended to be older than the healthy subjects, which could partially explain their lesser
response since FEVi responses to O3 exposure diminish with age. Individuals with
asthma also had more neutrophils in the BALF (18 hours postexposure) than
similarly exposed healthy individuals (Peden et al., 1997; Scannell et al., 1996;
Basha et al., 1994). Furthermore, a study examining the effects of O3 on individuals
with atopic asthma and healthy controls reported that greater numbers of neutrophils,
higher levels of cytokines and hyaluronan, and greater expression of macrophage
8-13
-------
cell-surface markers were observed in induced sputum of atopic asthmatics compared
with healthy controls (Hernandez et al., 2010). Differences in O3-induced epithelial
cytokine expression were noted in bronchial biopsy samples from asthmatics and
healthy controls (Bosson et al., 2003). Cell-surface marker and cytokine expression
results, and the presence of hyaluronan, are consistent with O3 having greater effects
on innate and adaptive immunity in these asthmatic individuals (see Section 5.4.2.2).
In addition, studies have demonstrated that O3 exposure leads to increased bronchial
reactivity to inhaled allergens in mild allergic asthmatics (Kehrl et al.. 1999: Jorres et
al.. 1996) and to the influx of eosinophils in individuals with pre-existing allergic
disease (Vagaggini et al.. 2002: Peden et al.. 1995). Taken together, these results
point to several mechanistic pathways which could account for the increased risk of
Os-related health effects in subjects with asthma (see Section 5.4.2.2).
Toxicological studies provide biological plausibility for greater effects of O3 among
those with asthma or AHR. In animal toxicological studies, an asthmatic phenotype
is modeled by allergic sensitization of the respiratory tract. Many of the studies that
provide evidence that O3 exposure is an inducer of AHR and remodeling utilize these
types of animal models. For example, a series of experiments in infant rhesus
monkeys have shown these effects, but only in monkeys sensitized to house dust mite
allergen (Fanucchi et al., 2006: Joad et al., 2006: Schelegle et al., 2003). Similarly,
Funabashi et al. (2004) demonstrated changes in pulmonary function in mice exposed
to O3, and Wagner et al. (2007) demonstrated enhanced inflammatory responses in
rats exposed to O3, but only in animals sensitized to allergen. In general, it is the
combined effects of O3 and allergic sensitization which result in measurable effects
on pulmonary function. In a bleomycin induced pulmonary fibrosis model, exposure
to 250 ppb O3 for 5 days increased pulmonary inflammation and fibrosis, along with
the frequency of bronchopneumonia in rats (Ovarzun et al.. 2005). Thus, short-term
exposure to O3 may enhance damage in a previously injured lung.
In the 2006 O3 AQCD, the potential for individuals with asthma to have greater risk
of O3-related health effects was supported by a number of controlled human
exposure studies, evidence from toxicological studies, and a limited number of
epidemiologic studies. Overall, in the recent epidemiologic literature some, but not
all, studies report greater risk of health effects among individuals with asthma.
Studies examining effect measure modification of the relationship between
short-term O3 exposure and altered lung function by corticosteroid use provided
limited and inconsistent evidence of O3-related health effects. Additionally, recent
studies of behavioral responses have found that studies do not take into account
individual behavioral adaptations to forecasted air pollution levels (such as avoidance
and reduced time outdoors), which may underestimate the observed associations in
studies that examined the effect of O3 exposure on respiratory health (Neidell and
Kinney, 2010). This could explain some inconsistency observed among recent
epidemiologic studies. The evidence from controlled human exposure studies
provides support for increased decrements in FEVi and greater inflammatory
responses to O3 in individuals with asthma than in healthy individuals without a
history of asthma. These studies are often performed among individuals with mild
asthma and therefore it is possible that individuals with severe asthma may have an
8-14
-------
even greater risk of O3-related health effects. The collective evidence for increased
risk of Os-related health effects among individuals with asthma from controlled
human exposure studies is supported by recent toxicological studies which provide
biological plausibility for heightened risk of asthmatics to respiratory effects due to
O3 exposure. Evidence indicating O3-induced respiratory effects among individuals
with asthma is further supported by additional studies of O3-related respiratory
effects (Section 6.2). Overall, there is adequate evidence for asthmatics to be an at-
risk population based on the substantial, consistent evidence among controlled
human exposure studies and coherence from epidemiologic and toxicological studies.
8.2.3 Chronic Obstructive Pulmonary Disease (COPD)
In the U.S. over 4% of adults report having chronic bronchitis and almost 2% report
having emphysema, both of which are classified as COPD (Pleis et al., 2009).
A recent study reported no association between O3 exposure and lung function
regardless of whether the study participant had COPD or other pre-existing diseases
(asthma or IHD) (Lagorio et al.. 2006).
Peel et al. (2007) found that individuals with COPD were at increased risk of
cardiovascular ED visits in response to short-term O3 exposure compared to healthy
individuals in Atlanta, GA. The authors reported that short-term O3 exposure was
associated with higher odds of an emergency department (ED) visit for peripheral
and cerebrovascular disease among individuals with COPD compared to individuals
without COPD. However, pre-existing COPD did not increase the odds of
hospitalization for all CVD outcomes (i.e., IHD, dysrhythmia, or congestive heart
failure). In an additional study performed in Taiwan, individuals with and without
COPD had higher odds of congestive heart failure associated with O3 exposure on
warm days (Lee et al.. 2008a). As discussed in Section 6.3. most studies reported no
overall association between O3 concentration and CVD morbidity.
Recent epidemiologic evidence indicates that persons with COPD may have
increased risk of O3-related cardiovascular effects, but little information is available
on whether COPD leads to an increased risk of O3-induced respiratory effects.
Overall, this small number of studies provides inadequate evidence to determine
whether COPD results in increased risk of O3-related health effects.
8.2.4 Cardiovascular Disease (CVD)
Cardiovascular disease has become increasingly prevalent in the U.S., with about
12% of adults reporting a diagnosis of heart disease (Table 8-4). A high prevalence
of other cardiovascular-related conditions has also been observed, such as
hypertension which is prevalent among approximately 24% of adults. In the 2006 O3
AQCD, little evidence was available regarding whether pre-existing CVD
8-15
-------
contributed to increased risk of O3-related health effects. Recent epidemiologic
studies have examined cardiovascular-related diseases as modifiers of the
O3-outcome associations; however, no recent evidence is available from controlled
human exposure studies or toxicological studies.
Peel et al. (2007) compared the associations between short-term O3 exposure and
cardiovascular ED visits in Atlanta, GA among multiple comorbid conditions.
The authors found no evidence of increased risk of cardiovascular ED visits in
individuals previously diagnosed with dysrhythmia, congestive heart failure, or
hypertension compared to healthy individuals. Similarly, a study in France examined
the association between O3 concentrations and ischemic cerebrovascular events
(ICVE) and myocardial infarction (MI) and the influence of multiple vascular risk
factors on any observed associations (Henrotin et al., 2010). The association between
O3 exposure and ICVE was elevated for individuals with multiple risk factors,
specifically individuals with diabetes or hypertension. For the association between
O3 and MI, increased odds were apparent only for those with hypercholesterolemia.
In a study conducted in Taiwan, a positive association was observed for O3 on warm
days and congestive heart failure hospital admissions, but the association did not
differ between individuals with/without hypertension or with/without dysrhythmia
(Lee et al., 2008a). Another study in Taiwan reported that the association between O3
levels and ED visits for arrhythmias were greater on warm days among those with
congestive heart failure compared to those without congestive heart failure; however,
the estimate and 95% CIs for those without congestive heart failure is completely
contained within the 95% CI of those with congestive heart failure (Chiu and Yang,
2009).
Although not studied extensively, a study has examined the increased risk of
O3-related changes in blood markers for individuals with CVD. There was a greater
association between O3 exposure and some, but not all, blood inflammatory markers
among individuals with a history of CVD (Liao et al., 2005). Liao et al. (2005) found
that increased fibrinogen was positively associated with short-term O3 exposure but
this association was present only among individuals with a history of CVD.
No association was observed among those without a history of CVD. However, for
another biomarker (vWF), CVD status did not modify the positive association with
short-term O3 exposure (Liao et al., 2005).
Mortality studies provide some evidence for a potential increase in O3-induced
mortality in individuals with pre-existing atrial fibrillation and atherosclerosis. In a
study of 48 U.S. cities, increased risk of mortality with short-term O3 exposure was
observed only among individuals with secondary atrial fibrillation (Medina-Ramon
and Schwartz. 2008). No association was observed for short-term O3 exposure and
mortality in a study of individuals with diabetes with or without CVD prior to death;
however, there was some evidence of increased risk of mortality during the warm
season if individuals had diabetes and atherosclerosis compared to only having
diabetes (Goldberg et al.. 2006).
Finally, although not extensively examined, a study explored whether a pre-existing
CVD increased the risk of an O3-induced respiratory effect. Lagorio et al. (2006)
8-16
-------
examined the effect of O3 exposure on lung function among participants with a
variety of pre-existing diseases, including IHD. No association was observed
regardless of whether the participant had IHD.
Overall, most short-term exposure studies did not report increased O3-related
cardiovascular morbidity for individuals with pre-existing CVD. However, as
discussed in Section 6.3, most studies reported no overall association between O3
concentration and CV morbidity. Thus, it is likely the association would be null
regardless of the stratification. A limited number of studies examined whether
cardiovascular disease modifies the association between O3 and respiratory effects.
There was some evidence that cardiovascular disease increases the risk of O3-related
mortality but again the number of studies was limited. Currently, evidence is
inadequate to classify pre-existing CVD as a potential at-risk factor for O3-related
health effects. Future research among those with CVD compared to those without
will increase the understanding of potential increased risk of O3-related health effects
among this group.
8.2.5 Diabetes
The literature has not extensively examined whether individuals with diabetes (about
8% of U.S. adults) are potentially at increased risk of O3-related health effects. In a
study of short-term O3 exposure and cardiovascular ED visits in Atlanta, GA, no
association was observed for individuals with or without diabetes (Peel et al.. 2007).
A similar study conducted in Taiwan reported a positive association between O3
exposure on warm days and hospital admissions for congestive heart failure;
however, no modification of the association by diabetes was observed (Lee et al..
2008a). Finally, in a study of O3 exposure and ED visits for arrhythmia in Taiwan,
there was no evidence of effect measure modification by diabetes on warm or cool
days (Chiu and Yang, 2009). Currently, the limited number of epidemiologic studies
as well as the lack of controlled human exposure or toxicological studies provides
inadequate evidence to indicate whether diabetes results in a potentially increased
risk of O3-related health effects.
8.2.6 Hyperthyroidism
Hyperthyroidism has been identified in toxicological studies as a potential factor that
may lead to increased risk of O3-related health effects but has not yet been explored
in epidemiologic or controlled human exposure studies. Lung damage and
inflammation due to oxidative stress may be modulated by thyroid hormones.
Compared to controls, hyperthyroid rats exhibited elevated levels of BAL neutrophils
and albumin after a 4-hour exposure to O3, indicating O3-induced inflammation and
damage. Hyperthyroidism did not affect production of reactive oxygen or nitrogen
species, but BAL phospholipids were increased, indicating greater activation of Type
II cells and surfactant protein production compared to normal rats (Huffman et al..
8-17
-------
2006). Thus, this study provides some underlying evidence, which suggests that
individuals with hyperthyroidism may represent an at-risk population; however,
overall the lack of additional studies provides inadequate evidence to determine
whether hyperthyroidism results in potentially increased risk of O3-related health
effects.
8.3 Sociodemographic Factors
8.3.1 Lifestage
The 1996 and 2006 O3 AQCDs identified children, especially those with asthma, and
older adults as at-risk populations. These previous AQCDs reported clinical
(controlled human exposure) evidence that children have greater spirometric
responses to O3 than middle-aged and older adults (U.S. EPA. 1996a). Similar results
were observed for symptomatic responses and O3 exposure. Among older adults,
most studies reported in the 2006 O3 AQCD reported greater effects of short-term O3
exposure and mortality compared to other age groups (U.S. EPA. 2006b). Evidence
published since the 2006 O3 AQCD, summarized below, further supports these
findings.
8.3.1.1 Children
The 2000 Census reported that 28.6% of the U.S. population was under 20 years of
age, with 14.1% under the age of 10 (SSDAN CensusScope. 2010a). Children's
respiratory systems are undergoing lung growth until about 18-20 years of age and
are therefore thought to be intrinsically more at risk for O3-induced damage (U.S.
EPA. 2006b). It is generally recognized that children spend more time outdoors than
adults, and therefore would be expected to have higher exposure to O3 than adults.
The ventilation rates also vary between children and adults, particularly during
moderate/heavy activity. Children aged 11 years and older and adults have higher
absolute ventilation rates than children aged 1-11 years. However, children have
higher ventilation rates relative to their lung volumes, which tends to increase dose
normalized to lung surface area. Exercise intensity has a substantial effect on
ventilation rate, with high intensity activities resulting in nearly double the
ventilation rate during moderate activity among children and those adults less than 31
years of age. For more information on time spent outdoors and ventilation rate
differences by age group, see Section 4.4.1.
The 1996 O3 AQCD, reported clinical evidence that children, adolescents, and young
adults (<18 years of age) appear, on average, to have nearly equivalent spirometric
responses to O3 exposure, but have greater responses than middle-aged and older
adults (U.S. EPA. 1996a). Symptomatic responses (e.g., cough, shortness of breath,
8-18
-------
pain on deep inspiration) to O3 exposure, however, appear to increase with age until
early adulthood and then gradually decrease with increasing age (U.S. EPA, 1996a).
For subjects aged 18-36 years, McDonnell et al. (1999b) reported that symptom
responses from O3 exposure also decrease with increasing age. Complete lung
growth and development is not achieved until 18-20 years of age in women and the
early 20s for men; pulmonary function is at its maximum during this time as well.
Additionally, PBPK modeling reported lung regional extraction of O3 to be higher in
infants compared to adults. This is thought to be due to the smaller nasal and
pulmonary regions' surface area in children under the age of 5 years compared to the
total airway surface area observed in adults (Sarangapani et al.. 2003).
Recent epidemiologic studies have examined different age groups and their risk to
O3-related respiratory hospital admissions and ED visits. A study in Cyprus of
short-term O3 concentrations and respiratory hospital admissions detected possible
effect measure modification by age with a larger association among individuals
<15 years of age compared with those >15 years of age. However, this difference
was only apparent with a 2-day lag (Middleton et al., 2008). Similarly, a Canadian
study of asthma-ED visits reported the strongest O3-related associations among 5 to
14 year-olds compared to the other age groups (ages examined 0-75+) (Villeneuve et
al., 2007). Greater O3-associated risk in asthma-related ED visits were also reported
among children (<15 years) as compared to adults (15 to 64 years) in a study from
Finland (Halonen et al., 2009). A study of New York City hospital admissions
demonstrated an increase in the association between O3 exposure and asthma-related
hospital admissions for 6 to 18 year-olds compared to those <6 years old and those
>18 years old (Silverman and Ito, 2010). When examining long-term O3 exposure
and asthma hospital admissions among children, associations were determined to be
larger among children 1 to 2 years old compared to children 2 to 6 years old (Lin et
al.. 2008b). A few studies reported positive associations among both children and
adults and no modification of the effect by age. A study performed in Hong Kong
examined O3 exposure and asthma-related hospital admissions for ages 0 to 14, 15 to
65, and >65 (Ko et al.. 2007). The researchers reported that the association was
greater among the 0 to 14 and 14 to 65 age groups compared to the >65 age group.
Another study looking at asthma-related ED visits and O3 exposure in Maine
reported positive associations for all age groups (ages 2 to 65) (Paulu and Smith.
2008). Effects of O3 exposure on asthma hospitalizations among both children and
adults (<18 and > 18 years old) were demonstrated in a study in Washington, but
only children (<18 years of age) had statistically significant results at lag day 0,
which the authors wrote, "suggests that children are more immediately responsive to
adverse effects of O3 exposure" (Mar and Koenig. 2009).
The evidence reported in epidemiologic studies is supported by recent toxicological
studies which observed O3-induced health effects in immature animals. Early life
exposures of multiple species of laboratory animals, including infant monkeys,
resulted in changes in conducting airways at the cellular, functional, ultra-structural,
and morphological levels. Carey et al. (2007) conducted a study of O3 exposure in
infant rhesus macaques, whose respiratory tract closely resembles that of humans.
Monkeys were exposed either acutely for 5 days to 0.5 ppm O3, or episodically for 5
8-19
-------
biweekly cycles alternating 5 days of 0.5 ppm O3 with 9 days of filtered air, designed
to mimic human exposure (70 days total). All monkeys acutely exposed to O3 had
moderate to marked necrotizing rhinitis, with focal regions of epithelial exfoliation,
numerous infiltrating neutrophils, and some eosinophils. The distribution, character,
and severity of lesions in episodically exposed infant monkeys were similar to that of
acutely exposed animals. Neither exposure protocol for the infant monkeys produced
mucous cell metaplasia proximal to the lesions, an adaptation observed in adult
monkeys exposed continuously to 0.3 ppm O3 in another study (Harkema et al..
1987a). Functional (increased airway resistance and responsiveness with antigen +
O3 co-exposure) and cellular changes in conducting airways (increased numbers of
inflammatory eosinophils) were common manifestations of exposure to O3 among
both the adult and infant monkeys (Plopper et al.. 2007). In addition, the lung
structure of the conducting airways in the infant monkeys was stunted by O3 and this
aberrant development was persistent 6 months postexposure. This developmental
endpoint was not, of course, studied in the adult monkey experiments (Fanucchi et
al.. 2006). Thus, some functional and biochemical effects were similar between the
infant and adult monkeys exposed to O3, but because the study designs did not
include concentration-response experiments, it is not possible to determine whether
the infant monkeys were more at risk for the effects of O3.
Similarly, rat fetuses exposed to O3 in utero had ultrastructural changes in
bronchiolar epithelium when examined near the end of gestation (Lopez et al., 2008).
In addition, exposure of mice to mixtures of air pollutants early in development
affected pup lung cytokine levels (TNF, IL-1, KC, IL-6, and MCP-1) (Auten et al.,
2009). In utero exposure of animals to PM augmented O3-induced airway
hyper-reactivity in these pups as juveniles.
Age may affect the inflammatory response to O3 exposure. In comparing neonatal
mice to adult mice, increased bronchoalveolar lavage (BAL) neutrophils were
observed in four strains of neonates 24 hours after exposure to 0.8 ppm O3 for
5 hours (Vancza et al., 2009). Three of these strains also exhibited increased BAL
protein, although the two endpoints were not necessarily consistently correlated in a
given strain. In some strains, however, adults were responsive, indicating a strain-age
interaction. Measurement of 18O determined that the observed strain- and age-
dependent differences were not due to absorbed O3 dose. Using electron microscopy,
Bils (1970) studied the lungs of mice of different ages (4 days or 1 to 2 months)
exposed to 0.6 to 1.3 ppm O3 for 6 to 7 h/day for 1 to 2 days and noted swelling of
the alveolar epithelial lining cells without intra-alveolar edema. Swelling of
endothelial cells and occasional breaks in the basement membrane were observed.
These effects were most evident in younger mice exposed for 2 days. Toxicological
studies reported that the difference in effects among younger lifestage test animals
may be due to age-related changes in endogenous antioxidants and sensitivity to
oxidative stress. A recent study demonstrated that 0.25 ppm O3 exposure
differentially altered expression of metalloproteinases in the skin of young (8 weeks
old) and aged (18 months old) mice, indicating age-related variations in risk of
oxidative stress (Fortino et al., 2007). Valacchi et al. (2007) found that aged mice had
more Vitamin E in their plasma but less in their lungs compared to young mice,
8-20
-------
which may affect their pulmonary antioxidant defenses. Servais et al. (2005) found
higher levels of oxidative damage indicators in immature (3 weeks old) and aged
(20 months old) rats compared to adult rats, the latter which were relatively resistant
to an intermittent 7-day exposure to 0.5 ppm O3. Immature rats exhibited a higher
ventilation rate, which may have increased exposure. Additionally, a series of
toxicological studies reported an association between O3 exposure and bradycardia
that was present among young but not older mice (Hamade et al.. 2010: Tankerslev et
al.. 2010: Hamade and Tankerslev. 2009: Hamade et al.. 2008). Regression analysis
revealed an interaction between age and strain on heart rate, which implies that aging
may affect heart rate differently among mouse strains (Tankerslev et al.. 2010).
The authors proposed that the genetic differences between the mice strains could be
altering the formation of ROS, which tends to increase with age, thus modulating the
changes in cardiopulmonary physiology after O3 exposure.
The previous and recent human clinical (controlled human exposure) and
toxicological studies reported evidence of increased risk from O3 exposure for
younger ages, which provides coherence and biological plausibility for the findings
from epidemiologic studies. Although there was some inconsistency, generally, the
epidemiologic studies reported larger associations for respiratory hospital admissions
and ED visits for children than adults. The interpretation of these studies is limited
by the lack of consistency in comparison age groups and outcomes examined.
Toxicological studies observed O3-induced health effects in immature animals,
including infant monkeys, though the effects were not consistently greater in young
animals than adults. However, overall, the epidemiologic, controlled human
exposure, and toxicological studies provide substantial and consistent evidence
within and across disciplines. Therefore, there is adequate evidence to conclude that
children are at increased risk of O3-related health effects.
8.3.1.2 Older Adults
Older adults may be at greater risk of health effects associated with O3 exposure
through a variety of intrinsic pathways. In addition, older adults may differ in their
exposure and internal dose. Older adults spend slightly more time outdoors than
adults aged 18-64 years. Older adults also have somewhat lower ventilation rates
than adults aged 31 - less than 61 years. For more information on time spent outdoors
and ventilation rate differences by age group, see Section 4.4.1. The gradual decline
in physiological processes that occur with aging may lead to increased risk of
O3-related health effects (U.S. EPA, 2006a). Respiratory symptom responses to O3
exposure appears to increase with age until early adulthood and then gradually
decrease with increasing age (U.S. EPA, 1996a), which may put older adults at
increased risk by withstanding continued O3 exposure and thus not seeking relief and
avoiding exposure. In addition, older adults, in general, have a higher prevalence of
pre-existing diseases, with the exception of asthma, compared to younger age groups
and this may also lead to increased risk of O3-related health effects (see Table 8-4
that gives pre-existing rates by age). With the number of older Americans increasing
8-21
-------
in upcoming years (estimated to increase from 12.4% of the U.S. population to
19.7% between 2000 to 2030, which is approximately 35 million and 71.5 million
individuals, respectively) this group represents a large population potentially at risk
of O3-related health effects (SSDAN CensusScope. 2010a: U.S. Census Bureau.
2010).
The majority of recent studies reported greater effects of short-term O3 exposure and
mortality among older adults, which is consistent with the findings of the 2006 O3
AQCD. A study conducted in 48 cities across the U.S. reported larger effects among
adults > 65 years old compared to those <65 years (Medina-Ramon and Schwartz,
2008). Further investigation of this study population revealed a trend of O3-related
mortality risk that gets larger with increasing age starting at age 50 (Zanobetti and
Schwartz, 2008a). A study of 7 urban centers in Chile reported similar results, with
greater effects in adults > 65 years old, however the effects were smaller among
those > 85 years old compared to those in the 75 to 84 years old age range (Cakmak
et al., 2007). More recently, a study conducted in the same area reported similar
associations between O3 exposure and mortality in adults aged <64 years old and 65
to 74 years old, but the risk was increased among older age groups (Cakmak et al.,
2011). A study performed in China reported greater effects in populations > 45 years
old (compared to 5 to 44 year-olds), with statistically significant effects present only
among those > 65 years old (Kan et al., 2008). An Italian study reported higher risk
of all-cause mortality associated with increased O3 concentrations among individuals
> 85 years old as compared to those 35 to 84 years old. Those 65 to 74 and 75 to
84 years old did not show a greater increase in risk compared to those aged 35 to
64 years (Stafoggia et al., 2010). The Air Pollution and Health: A European and
North American Approach (APHENA) project examined the association between O3
exposure and mortality for those <75 and > 75 years of age. In Canada, the
associations for all-cause and cardiovascular mortality were greater among those
> 75 years old in the summer-only and all-year analyses. Age groups were not
compared in the analysis for respiratory mortality in Canada. In the U.S., the
association for all-cause mortality was slightly greater for those <75 years of age
compared to those > 75 years old in summer-only analyses. No consistent pattern
was observed for CVD mortality. In Europe, slightly larger associations for all-cause
mortality were observed in those <75 years old in all-year and summer-only
analyses. Larger associations were reported among those <75years for CVD
mortality in all-year analyses, but the reverse was true for summer-only analyses
(Katsouyanni et al.. 2009).
Multiple epidemiologic studies of O3 exposure and hospital admissions were
stratified by age groups. A positive association was reported between short-term O3
exposure and respiratory hospital admissions for adults > 65 years old but not for
those adults aged 15 to 64 years (Halonen et al., 2009). In the same study, no
association was observed between O3 concentration and respiratory mortality among
those > 65 years old or those 15 to 64 years old; however, an inverse association
between O3 concentration and cardiovascular mortality was present among
individuals > 65 years old but not among individuals <65 years old. This inverse
association among those > 65 years old persisted when examining hospital
8-22
-------
admissions for coronary heart disease. A study of CVD-related hospital visits in
Bangkok, Thailand reported an increase in percent change for hospital visits with
previous day and cumulative 2-day O3 levels among those > 65 years old, whereas
no association was present for individuals less than 65 years of age (Buadong et al..
2009). No association was observed for current day or cumulative 3-day averages in
any age group. A study examining O3 and hospital admissions for CVD-related
health effects reported no association for individuals aged 15 to 64 or individuals
aged > 65 years, although one lag-time did show an inverse effect for coronary heart
disease among elderly that was not present among 15 to 64 year-olds (Halonen et al..
2009). However, as discussed in the section on CVD hospital admissions
(Section 6.3.2.7). results were inconsistent and often null so it is plausible that no
association would be observed regardless of age. No modification by age (40 to
64 year-olds versus >64 years old) was observed in a study from Brazil examining
O3 levels and COPD ED visits (Arbex et al.. 2009).
Biological plausibility for differences by age is provided by toxicological studies.
Ozone exposure resulted in an increase in left ventricular chamber dimensions at end
diastole (LVEDD) in young and old mice, whereas decreases in left ventricular
posterior wall thickness at end systole (PWTES) were only observed among older
mice (Tankersley et al.. 2010). Other toxicological studies also indicate increased
risk in older animals for additional endpoints, including neurological and immune.
The hippocampus, one of the main regions affected by age-related neurodegenerative
diseases, may be more sensitive to oxidative damage in aged rats. In a study of young
(47 days) and aged (900 days) rats exposed to 1 ppm O3 for 4 hours, O3-induced
lipid peroxidation occurred to a greater extent in the striatum of young rats, whereas
it was highest in the hippocampus in aged rats (Rivas-Arancibia et al.. 2000).
In young mice, healing of skin wounds is not significantly affected by O3 exposure
(Lim et al.. 2006). However, exposure to 0.5 ppm O3 for 6 h/day significantly delays
wound closure in aged mice.
Although some outcomes reported mixed findings regarding an increase in risk for
older adults, recent epidemiologic studies report consistent positive associations
between short-term O3 exposure and mortality in older adults. The evidence from
mortality studies is consistent with the results reported in the 2006 O3 AQCD and is
supported by toxicological studies providing biological plausibility for increased risk
of effects in older adults. Also, older adults may be experiencing increased exposure
compared to younger adults due to time spend outdoors and withstanding exposures.
Overall, adequate evidence is available indicating that older adults are at increased
risk of O3-related health effects based on the substantial and consistent evidence
within epidemiologic studies on O3 exposure and mortality and the coherence with
toxicological studies.
8-23
-------
8.3.2 Sex
The distribution of males and females in the U.S. is similar. In 2000, 49.1% of the
U.S. population was male and 50.9% were female. However, this distribution does
vary by age with a greater prevalence of females > 65 years old compared to males
(SSDAN CensusScope. 2010a). The 2006 O3 AQCD did not report evidence of
differences between the sexes in health responses to O3 exposure (U.S. EPA, 2006b).
Recent epidemiologic studies have evaluated the effects of short-term and long-term
exposure to O3 on multiple health endpoints stratified by sex.
A study in Maine that examined short-term O3 concentrations and asthma ED visits
detected greater effects among males ages 2 to 14 years and among females ages 15
to 34 years compared to males and females in the same age groups (no difference
was detected for males and females aged 35 to 64) (Paulu and Smith. 2008).
A Canadian study reported no associations between short-term O3 and respiratory
infection hospital admissions for either boys or girls under the age of 15 (Lin et al..
2005). whereas another Canadian study reported a slightly higher but
non-statistically significant increase in respiratory hospital admissions for males
(mean ages 47.6 to 69.0 years) (Cakmak et al.. 2006b). A recent study from Hong
Kong examining individuals of all ages reported no effect measure modification by
sex for overall respiratory disease hospital admissions, but did detect a greater excess
risk of hospital admissions for COPD among females compared to males (Wong et
al.. 2009). Similarly a study in Brazil found higher effect estimates for COPD ED
visits among females compared to males (Arbex et al.. 2009). Higher levels of
respiratory hospital admissions with greater O3 concentrations was also observed for
females in a study of individuals living in Cyprus (Middleton et al., 2008). A study of
lung function unrelated to hospital admissions and ED visits was conducted among
lifeguards in Texas and reported decreased lung function with increased O3 exposure
among females but not males (Thaller et al., 2008). This study included individuals
aged 16 to 27 years, and the majority of participants were male. A New York study
found no evidence of effect measure modification of the association between
long-term O3 exposure and asthma hospital admissions among males and females
between 1 and 6 years old (Lin et al., 2008b).
In addition to examining the potential modification of O3 associations with
respiratory outcomes by sex, studies also examined cardiovascular-related outcomes
specifically hospital admissions and ED visits. All of these studies reported no effect
modification by sex with some studies reporting null associations for both males and
females (Wong et al.. 2009: Middleton et al.. 2008: Villeneuve et al.. 2006a) and one
study reporting a positive associations for both sexes (Cakmak et al.. 2006a).
A French study examining the associations between O3 concentrations and risk of
ischemic strokes (not limited to ED visits or hospital admissions) reported no
association for either males or females with lags of 0, 2, or 3 days (Henrotin et al..
2007). A positive association was reported for males with a lag of 1 day, but this
association was null for females. The authors noted that men in the study had much
higher rates of current and former smoking than women (67.4% versus 9.3%).
Additionally, cardiovascular hospital admissions and ED visits overall have
8-24
-------
demonstrated inconsistent and null results (Section 6.3.2.7). The lack of effect
measure modification by sex may be indicative of the lack of association, not the
lack of an effect by sex.
A biomarker study investigating the effects of O3 concentrations on high-sensitivity
C-reactive protein (hs-CRP), fibrinogen, and white blood cell (WBC) count, reported
observations for various lag times ranging from 0 to 7 days (Steinvil et al.. 2008).
Most of the associations were null for males and females although one association
between O3 and fibrinogen was positive for males and null for females (lag day 4);
however, this positive association was null or negative when other pollutants were
included in the model. One study examining correlations between O3 levels and
oxidative DNA damage examined results stratified by sex. In this study Palli et al.
(2009) reported stronger correlations for males than females, both during short-term
exposure (less than 30 days) and long-term exposure (0-90 days). However, the
authors commented that this difference could have been partially explained by
different distributions of exposure to traffic pollution at work.
A few studies have examined the association between short-term O3 concentrations
and mortality stratified by sex and, in contrast with studies of other endpoints, were
more consistent in reporting elevated risks among females. These studies, conducted
in the U.S. (Medina-Ramon and Schwartz. 2008). Italy (Stafoggia et al.. 2010). and
Asia (Kan et al.. 2008). reported larger effect estimates in females compared to
males. In the U.S. study, the elevated risk of mortality among females was greater
specifically among those > 60 years old (Medina-Ramon and Schwartz. 2008).
However, a recent study in Chile reported similar associations between O3 exposure
and mortality among both men and women (Cakmak et al.. 2011). A long-term O3
exposure study of respiratory mortality stratified their results by sex and reported
relative risks of 1.01 (95% CI: 0.99, 1.04) for males and 1.04 (95% CIs 1.03, 1.07)
for females (Jerrett et al.. 2009).
Experimental research provided a further understanding of the underlying
mechanisms that may explain a possible differential risk in O3-related health effects
among males and females. Several studies have suggested that physiological
differences between sexes may predispose females to greater effects from O3.
In females, lower plasma and nasal lavage fluid (NLF) levels of uric acid (most
prevalent antioxidant), the initial defense mechanism of O3 neutralization, may be a
contributing factor (Houslev et al.. 1996). Consequently, reduced absorption of O3 in
the upper airways of females may promote its deeper penetration. Dosimetric
measurements have shown that the absorption distribution of O3 is independent of
sex when absorption is normalized to anatomical dead space (Bush et al.. 1996).
Thus, a differential removal of O3 by uric acid seems to be minimal. In general, the
physiologic response of young healthy females to O3 exposure appears comparable
to the response of young males (Hazucha et al.. 2003). A few studies have examined
changes in O3 responses during various menstrual cycle phases. Lung function
response to O3 was enhanced during the follicular phase of the menstrual cycle
compared to the luteal phase in a small study of women (Fox et al.. 1993). However,
Seal et al. (1996) later reported no effect of menstrual cycle phase in their analysis of
8-25
-------
responses from 150 women, but conceded that the methods used by Fox et al. (1993)
more precisely defined the menstrual cycle phase. Another study also reported no
difference in responses among females during the follicular and luteal phases of their
cycle (Weinmann et al., 1995c). Additionally, in this study the responses in women
were comparable to those reported for men in the study. In a toxicological study,
small differences in effects by sex were seen in adult mice with respect to pulmonary
inflammation and injury after a 5-h exposure to 0.8 ppm O3, and although adult
females were generally more at risk, these differences were strain-dependent, with
some strains exhibiting greater risk in males (Vancza et al.. 2009). The most obvious
sex difference was apparent in lactating females, which incurred the greatest lung
injury or inflammation among several of the strains.
Overall, results have varied, with recent evidence for increased risk for O3-related
health effects present for females in some studies and males in other studies. Most
studies examining the associations O3 and mortality report females to be at greater
risk than males, but minimal evidence is available regarding a difference between the
sexes for other outcomes. Inconsistent findings were reported on whether effect
measure modification exists by sex for respiratory and cardiovascular hospital
admissions and ED visits, although there is some indication that females are at
increased risk of O3-related respiratory hospital admissions and ED visits. While O3-
related effects may occur in both men and women, there is suggestive evidence exists
indicating that females are at potentially increased risk of O3-related health effects as
there are consistent findings among epidemiologic studies of mortality.
8.3.3 Socioeconomic Status
SES is often represented by personal or neighborhood SES, which is comprised of a
variety of components such as educational attainment, household income, health
insurance status, and other such factors. SES is often indicative of such things as
access to healthcare, quality of housing, and pollution gradient to which people are
exposed. One or a combination of these components could modify the risk of O3-
related health effects. Based on the 2000 Census data, 12.4% of Americans live in
poverty (poverty threshold for a family of four was $17,463) (SSDAN CensusScope,
2010c). Although included below, studies stratifying by SES that are conducted
outside the U.S. may not be comparable to those studies from within the United
States. Having low SES in another country may be different than having low SES in
the U.S. based on SES definitions, population composition, and/or conditions in that
country.
Multiple epidemiologic studies have reported individuals of low SES to have
increased risk for the effects of short-term O3 exposure on respiratory hospital
admissions and ED visits. In New York State, larger associations between long-term
O3 exposure and asthma hospital admissions were observed among children of
mothers who did not graduate from high school, whose births were covered by
Medicaid/self-paid, or who were living in poor neighborhoods compared to children
8-26
-------
whose mothers graduated from high school, whose births were covered by other
insurance, or who were not living in poor neighborhoods, respectively (Lin et al.,
2008b). In addition, a study conducted across 10 cities in Canada found the largest
association between O3 exposure and respiratory hospital admissions was among
those with an educational level less than grade 9, but no consistent trend in the effect
was seen across quartiles of income (Cakmak et al.. 2006b). A Canadian study
reported inverse effects of O3 on respiratory hospital admissions and ED visits for all
levels of SES, measured by average census tract household income (Burraet al..
2009). A study performed in Korea examined the association between O3
concentrations and asthma hospital admissions and reported larger effect estimates in
areas of moderate and low SES compared with areas of high SES (SES was based on
average regional insurance rates) (Lee et al.. 2006).
The examination of the potential effects of SES on O3-related cardiovascular health
effects is relatively limited. A study conducted in Canada reported the association
between short-term O3 and ED visits for cardiac disease by quartiles of
neighborhood-level education and income. No effect measure modification was
apparent for either measure of SES (Cakmak et al.. 2006a). However, this may be
due to the lack of association present between O3 and ED visits for cardiac disease
regardless of SES.
Several studies were conducted that examined the modification of the relationship
between short-term O3 concentrations and mortality by SES. A U.S. multicity study
reported that communities with a higher proportion of the population unemployed
had higher O3-related mortality effect estimates (Bell and Dominici. 2008). A study
in seven urban centers in Chile reported on modification of the association between
O3 exposure and mortality using multiple SES markers (Cakmak et al.. 2011).
Increased risk was observed among the categories of low SES for all measures
(personal educational attainment, personal occupation, community income level).
Additionally, the APHENA study, which examined the association between O3 and
mortality by percentage unemployed, reported a higher percent change in mortality
with increased percent unemployed but this varied across the regions included in the
study (U.S., Canada, Europe) (Katsouyanni et al.. 2009). A Chinese study reported
that the greatest effects between O3 concentrations and mortality at lag day 0 were
among individuals living in areas of high social deprivation (i.e., low SES), but this
association was not consistent across lag days (at other lag times, the middle social
deprivation index category had the greatest association) (Wong et al.. 2008).
However, another study in Asia comparing low to high educational attainment
populations reported no evidence of greater mortality effects (total, CVD, or
respiratory) (Kan et al.. 2008). Additionally, a study in Italy reported no difference in
risk of mortality among census-block level derived income levels (Stafoggiaet al..
2010). A study of infant mortality in Mexico reported no association between O3
concentrations and infant mortality among any of the three levels of SES determined
using a socioeconomic index based on residential areas (Romieu et al.. 2004a).
Another study in Mexico reported a positive association between O3 levels at lag 0
and respiratory-related infant mortality in only the low SES group (determined based
on education, income, and household conditions across residential areas), but no
8-27
-------
association was observed in any of the SES groups with other lags (Carbajal-Arroyo
etal..20in.
Studies of O3 concentrations and reproductive outcomes have also examined
associations by SES levels. A study in California reported greater decreases in birth
weight associated with full pregnancy O3 concentration for those with neighborhood
poverty levels of at least 7% compared with those in neighborhoods with less than
7% poverty (the authors do not provide information on how categories of the SES
variable were determined) (Morello-Frosch et al., 2010). No dose response was
apparent and those with neighborhood poverty levels of'7-21% had greater decreases
observed for the association than those living in areas with poverty rates of at least
22%. An Australian study reported an inverse association between O3 exposure
during days 31-60 of gestation and abdominal circumference during gestation
(Hansen et al., 2008). The interaction with SES (area-level measured socioeconomic
disadvantage) was examined and although the inverse association remained
statistically significant in only the highest SES quartile, there were large confidence
interval overlaps among estimates for each quartile so no difference in the
association for the quartiles was apparent.
Evidence from a controlled human exposure study that examined O3 effects on lung
function does not provide support for greater O3-related health effects in individuals
of lower SES. In a follow-up study on modification by race, Seal et al. (1996)
reported that, of three SES categories, individuals in the middle SES category
showed greater concentration-dependent decline in percent-predicted FEVi (4-5% at
400 ppb O3) than in low and high SES groups. The authors did not have an
"immediately clear" explanation for this finding and controlled human exposure
studies are typically not designed to answer questions about SES.
Overall, most studies of individuals have reported that individuals with low SES and
those living in neighborhoods with low SES are more at risk for O3-related health
effects, resulting in increased risk of respiratory hospital admissions and ED visits.
Inconsistent results have been observed in the few studies examining effect
modification of associations between O3 exposure and mortality and reproductive
outcomes. Also, a controlled human exposure study does not support evidence of
increased risk of respiratory morbidity among individuals with lower SES. Overall,
evidence is suggestive of SES as a factor affecting risk of O3-related health outcomes
based on collective evidence from epidemiologic studies of respiratory hospital
admissions but inconsistency among epidemiologic studies of mortality and
reproductive outcomes. Further studies are needed to confirm this relationship,
especially in populations within the U.S.
8.3.4 Race/Ethnicity
Based on the 2000 Census, 69.1% of the U.S. population identified as non-Hispanic
whites. Approximately 12.1% of people reported their race/ethnicity as non-Hispanic
black and 12.6% reported being Hispanic (SSDAN CensusScope, 2010b).
8-28
-------
Only a few studies examined the associations between short-term O3 concentrations
and mortality and reported higher effect estimates among blacks (Medina-Ramon and
Schwartz, 2008) and among communities with larger proportions of blacks (Bell and
Dominici, 2008). Another study examined long-term exposure to O3 concentrations
and asthma hospital admissions among children in New York State. These authors
reported no statistically significant difference in the odds of asthma hospital
admissions for blacks compared to other races but did detect higher odds for
Hispanics compared to non-Hispanics (Lin et al.. 2008b).
Additionally, recent epidemiologic studies have stratified by race when examining
the association between O3 concentration and birth outcomes. A study conducted in
Atlanta, GA reported decreases in birth weight with increased third trimester O3
concentrations among Hispanics but not among non-Hispanic whites (Darrow et al.,
201 Ib). A California study reported that the greatest decrease in birth weight
associated with full pregnancy O3 concentration was among non-Hispanic whites
(Morello-Frosch et al., 2010). This inverse association was also apparent, although
not as strong, for Hispanics and non-Hispanic blacks. Increased birth weight was
associated with higher O3 exposure among non-Hispanic Asians and Pacific
Islanders but these results were not statistically significant.
Similar to the epidemiologic studies, a controlled human exposure study suggested
differences in lung function responses by race (Seal et al.. 1993). The independent
effects of sex-race group and O3 concentration on lung function were positive, but
the interaction between sex-race group and O3 concentration was not statistically
significant. The findings indicated some overall difference between the sex-race
groups that was independent of O3 concentration (the concentration-response curves
for the four sex-race groups are parallel). In a multiple comparison procedure on data
collapsed across all O3 concentrations for each sex-race group, both black men and
black women had larger decrements in FEVi than did white men. The authors noted
that the O3 dose per unit of lung tissue would be greater in blacks and females than
whites and males, respectively. That this difference in tissue dose might have
affected responses to O3 cannot be ruled out. The college students recruited for the
Seal et al. (1993) study were probably from better educated and more SES
advantaged families, thus reducing potential for these variables to be confounding
factors. Que et al. (2011) also examined pulmonary responses to O3 exposure in
blacks of African American ancestry and in whites. On average, the black males
experienced the greatest decrements in FEVi following O3 exposure. This decrease
was larger than the decrement observed among black females, white males, and
white females.
Overall, the results of recent studies indicate that there may be race-related increase
in risk of O3-related health effects for some outcomes, although the overall
understanding of potential effect measure modification by race is limited by the small
number of studies. Additionally, these results may be confounded by other factors,
such as SES. Overall, evidence is inadequate to determine if O3-related health effects
vary by race because of the insufficient quantity of studies and lack of consistency
within disciplines.
8-29
-------
8.4 Behavioral and Other Factors
8.4.1 Diet
Diet was not examined as a factor potentially affecting risk in previous O3 AQCDs,
but recent studies have examined modification of the association between O3 and
health effects by dietary factors. Because O3 mediates some of its toxic effects
through oxidative stress, the antioxidant status of an individual is an important factor
that may contribute to increased risk of O3-related health effects. Supplementation
with Vitamins C and E has been investigated in a number of studies as a means of
inhibiting O3-mediated damage.
Epidemiologic studies have examined effect measure modification by diet and found
evidence that certain dietary components are related to the effect O3 has on
respiratory outcomes. In a recent study the effects of fruit/vegetable intake and
Mediterranean diet was examined (Romieu et al., 2009). Increases in these food
patterns, which have been noted for their high Vitamins C and E and omega-3 fatty
acid content, protected against O3-related decreases in lung function among children
living in Mexico City. Another study examined supplementation of the diets of
asthmatic children in Mexico with Vitamins C and E (Sienra-Monge et al., 2004).
Associations were detected between short-term O3 exposure and nasal airway
inflammation among children in the placebo group but not in those receiving the
supplementation. The authors concluded that "Vitamin C and E supplementation
above the minimum dietary requirement in asthmatic children with a low intake of
Vitamin E might provide some protection against the nasal acute inflammatory
response to ozone."
The epidemiologic evidence is supported by controlled human exposure studies,
which have shown that the first line of defense against oxidative stress is
antioxidants-rich extracellular lining fluid (ELF) which scavenge free radicals and
limit lipid peroxidation. Exposure to O3 depletes the antioxidant level in nasal ELF
probably due to scrubbing of O3 (Mudway et al., 1999a); however, the concentration
and the activity of antioxidant enzymes either in ELF or plasma do not appear to be
related to O3 responsiveness (e.g., pulmonary function and inflammation) (Samet et
al.,2001; Avissar et al., 2000; Blomberg et al., 1999). Carefully controlled studies of
dietary antioxidant supplementation have demonstrated some protective effects of
a-tocopherol (a form of Vitamin E) and ascorbate (Vitamin C) on spirometric
measures of lung function after O3 exposure but not on the intensity of subjective
symptoms and inflammatory response including cell recruitment, activation and a
release of mediators (Samet et al., 2001; Trenga et al., 2001). Dietary antioxidants
have also afforded partial protection to asthmatics by attenuating postexposure
bronchial hyperresponsiveness (Trenga et al.. 2001).
Toxicological studies provide evidence of biological plausibility to the epidemiologic
and controlled human exposure studies. Wagner et al. (2009); (2007) found
reductions in O3-exacerbated nasal allergy responses in rats with y-tocopherol
8-30
-------
treatment (a form of Vitamin E). O3-induced inflammation and mucus production
were also inhibited by y-tocopherol. Supplementation with Vitamins C and E
partially ameliorated inflammation, oxidative stress, and airway hyperresponsiveness
in guinea pigs exposed subchronically to 0.12 ppm O3 ppm (Chhabra et al., 2010).
Inconsistent results were observed in other toxicological studies of Vitamin C
deficiency and O3-induced responses. Guinea pigs deficient in Vitamin C displayed
only minimal injury and inflammation after exposure to O3 (Kodavanti et al.. 1995).
A recent study in mice demonstrated a protective effect of (3-carotene in the skin,
where it limited the production of proinflammatory markers and indicators of
oxidative stress induced by O3 exposure (Valacchi et al.. 2009). Deficiency of
Vitamin A, which has a role in regulating the maintenance and repair of the epithelial
layer, particularly in the lung, appears to enhance the risk of O3-induced lung injury
(Paquette et al.. 1996). Differentially susceptible mouse strains that were fed a
Vitamin A sufficient diet were observed to have different tissue concentrations of the
vitamin, potentially contributing to their respective differences in O3-related
outcomes. In addition to the studies of antioxidants, one toxicological study
examined protein deficiency. Protein deficiency alters the levels of enzymes and
chemicals in the brain of rats involved with redox status; exposure to 0.75 ppm O3
has been shown to differentially affect Na+/K+ ATPase, glutathione, and lipid
peroxidation, depending on the nutritional status of the animal, but the significance
of these changes is unclear (Calderon Guzman et al.. 2006). There may be a
protective effect of overall dietary restriction with respect to lung injury, possibly
related to increased Vitamin C in the lung surface fluid (Kari et al.. 1997).
There is adequate evidence that individuals with reduced intake of Vitamins E and C
are at risk for O3-related health effects based on substantial, consistent evidence both
within and among disciplines. The evidence from epidemiologic studies is supported
by controlled human exposure and toxicological studies.
8.4.2 Obesity
Obesity, defined as aBMI of 30 kg/m2 or greater, is an issue of increasing
importance in the U.S., with self-reported rates of obesity of 26.7% in 2009, up from
19.8% in 2000 (Sherry et al.. 2010). BMI may affect O3-related health effects
through multiple avenues, such as, inflammation in the body, increased pre-existing
disease, and poor diet. Increased risk of PM-related health effects have been
observed among obese individuals compared with non-obese individuals
[2009 PM ISA (U.S. EPA. 2009d)1
A few studies have been performed examining the association between BMI and
O3-related changes in lung function. An epidemiologic study reported decreased lung
function with increased short-term O3 exposure for both obese and non-obese
subjects; however, the magnitude of the reduction in lung function was greater for
those subjects who were obese (Alexeeff et al.. 2007). Further decrements in lung
function were noted for obese individuals with AHR. Controlled human exposure
8-31
-------
studies have also detected differential effects of O3 exposure on lung function for
individuals with varying BMIs. In a retrospective analysis of data from 541 healthy,
nonsmoking, white males between the ages of 18-35 years from 15 studies conducted
at the U.S. EPA Human Studies Facility in Chapel Hill, North Carolina, McDonnell
et al. (2010) found that increased body mass index (BMI) was found to be associated
with enhanced FEVi responses. The BMI effect was of the same order of magnitude
but in the opposite direction of the age effect whereby FEVi responses diminish with
increasing age. In a similar analysis, Bennett et al. (2007) found enhanced FEVi
decrements following O3 exposure with increasing BMI in a group of healthy,
nonsmoking, women (BMI range 15.7 to 33.4), but not among healthy, nonsmoking
men (BMI range 19.1 to 32.9). In the women, greater O3-induced FEVi decrements
were seen in individuals that were overweight/obese (BMI >25) compared to normal
weight (BMI from 18.5 to 25), and in normal weight compared to underweight (BMI
<18.5). Even disregarding the five underweight women, a greater O3 response in the
overweight/obese category (BMI >25) was observed compared with the normal
weight group (BMI from 18.5 to 24.9).
Studies in genetically and dietarily obese mice have shown enhanced pulmonary
inflammation and injury with acute O3 exposure (Johnston et al., 2008; Shore, 2007).
However, a recent study found that obese mice are actually resistant to O3-induced
pulmonary injury and inflammation and reduced lung compliance following longer
exposures (72 hours) at lower concentrations (0.3 ppm O3), regardless of whether
obesity was genetic- or diet-induced (Shore et al., 2009).
Multiple epidemiologic, controlled human exposure, and toxicological studies have
reported suggestive evidence for increased O3-related respiratory health effects
among obese individuals. Future research of the effect modification of the
relationship between O3 and other health-related outcomes besides respiratory health
effects by BMI and studies examining the role of physical conditioning will advance
understanding of obesity as a factor potentially increasing an individual's risk.
8.4.3 Smoking
Previous O3 AQCDs have concluded that smoking does not increase the risk of
O3-related health effects; in fact, in controlled human exposure studies, smokers have
been found to be less responsive to O3 than non-smokers. Data from recent
interviews conducted as part of the 2008 National Health Interview Survey (NHIS)
(Pleis et al.. 2009) have shown the rate of smoking among adults > 18 years old to be
approximately 20% in the United States. Approximately 21% of individuals surveyed
were identified as former smokers.
Baccarelli et al. (2007) performed a study of O3 concentrations and plasma
homocysteine levels (a risk factor for vascular disease). They found no interaction of
smoking (smokers versus non-smokers) for the associations between O3
concentrations and plasma homocysteine levels. Another study examined the
association between O3 and resting heart rate and also reported no interaction with
8-32
-------
smoking status (current smokers versus current non-smokers) (Ruidavets et al.,
2005a).
A study examining correlations between O3 levels and oxidative DNA damage
examined results stratified by current versus never and former smokers (Palli et al..
2009). Ozone was positively associated with DNA damage for short-term and
long-term exposures among never/former smokers. For current smokers, short-term
O3 concentrations were inversely associated with DNA damage; however, the
number of current smokers in the study was small (n = 12).
The findings of Palli et al. (2009) were consistent with those from controlled human
exposure studies that have confirmed that smokers are less responsive to O3 exposure
than non-smokers. Spirometric and plethysmographic pulmonary function decline,
nonspecific AHR, and inflammatory responses of smokers to O3 exposure were all
weaker than those reported for non-smokers. Similarly, the time course of
development and recovery from these effects, as well as their reproducibility, was not
different from non-smokers. Chronic airway inflammation with desensitization of
bronchial nerve endings and an increased production of mucus may plausibly explain
the pseudo-protective effect of smoking (Frampton et al.. 1997a: Torres et al.. 1997).
These findings for smoking are consistent with the conclusions from previous
AQCDs. An epidemiologic study of O3-associated DNA damage reported smokers to
be less at risk for O3-related health effects. In addition, both epidemiologic studies of
short-term exposure and CVD outcomes found no evidence of effect measure
modification by smoking. No toxicological studies provide biological support for O3-
related effects. Overall, evidence of potential differences in O3-related health effects
by smoking status is inadequate due to insufficient coherence and a limited number
of studies.
8.4.4 Outdoor Workers
Studies included in the 2006 O3 AQCD reported that individuals who participate in
outdoor activities or work outside to be a population at increased risk based on
consistently reported associations between O3 exposure and respiratory health
outcomes in these groups (U.S. EPA, 2006b). Outdoor workers are exposed to
ambient O3 concentrations for a greater period of time than individuals who spend
their days indoors. As discussed in Section 4.3.3 of this ISA, outdoor workers
sampled during the work shift had a higher ratio of personal exposure to fixed-site
monitor concentrations than health clinic workers who spent most of their time
indoors. Additionally, an increase in dose to the lower airways is possible during
outdoor exercise due to both increases in the amount of air breathed (i.e., minute
ventilation) and a shift from nasal to oronasal breathing (Sawyer et al.. 2007;
Nodelman andUltman, 1999; Hu et al.. 1994). For further discussion of the
association between FEVi responses to O3 exposure and minute ventilation, refer to
Section 6.2.3.1 of the 2006 O3 AQCD. A recent study has explored the potential
effect measure modification of O3 exposure and DNA damage by indoor/outdoor
8-33
-------
workplace (Tovalin et al., 2006). In a study of indoor and outdoor workers in
Mexico, individuals who worked outdoors in Mexico City had a slight association
between O3 exposure and DNA damage (measured by comet tail length assay),
whereas no association was observed for indoor workers. However, workers in
another Mexican city, Puebla, demonstrated no association between O3 levels and
DNA damage, regardless of whether they worked indoors or outdoors.
Previous studies have shown that increased exposure to O3 due to outdoor work leads
to increased risk of O3-related health effects, specifically decrements in lung function
(U.S. EPA, 2006b). Additionally, outdoor workers may be an at-risk population due
to their increased dose and exposure to O3. Recent evidence from a stratified analysis
does not indicate that increased O3 exposure due to outdoor work leads to DNA
damage. However, the strong evidence from the 2006 O3 AQCD which demonstrated
increased exposure, dose, and ultimately risk of O3-related health effects in this
population supports that there is adequate evidence available to indicate that
increased exposure to O3 through outdoor work increases the risk of O3-related
health effects.
8.4.5 Air Conditioning Use
Air conditioning use is an important indicator of O3 exposure, as use of central air
conditioning will limit exposure to O3 by blocking the penetration of O3 into the
indoor environment and lack of air conditioning may be linked to increased exposure
by use of open windows (see Section 4.3.2). Air conditioning use is a difficult effect
measure modifier to examine in epidemiologic studies because it is often estimated
using regional prevalence data and may not reflect individual-level use. More
generally, air conditioning prevalence is associated with temperature of a region;
those areas with higher temperatures have a greater prevalence of households with air
conditioning. Therefore, not having air conditioning is not necessarily indicative of
higher O3 exposure. Despite these limitations, a few studies have examined effect
measure modification by prevalence of air conditioning use in an area. Studies
examining multiple cities across the U.S. have assessed whether associations
between O3 concentrations and hospital admissions and mortality varied among areas
with high and low prevalence of air conditioning. Medina-Ramon et al. (2006)
conducted a study during the warm season and observed a greater association
between O3 levels and pneumonia-hospital admissions among areas with a lower
proportion of households having central air conditioning compared to areas with a
larger proportion of households with air conditioning. However, a similar
observation was not observed when examining COPD hospital admissions
complicating the interpretation of the results from this study. Bell and Dominici
(2008) found evidence of increased risk of O3-related mortality in areas with a lower
prevalence of central air conditioning in a study of 98 U.S. communities. Conversely,
Medina-Ramon and Schwartz (2008) found that among individuals with atrial
fibrillation, a lower risk of mortality was observed for areas with a lower prevalence
of central air conditioning.
8-34
-------
The limited number of studies that examined whether air conditioning use modifies
the association between O3 exposure and health has not provided consistent evidence
across health endpoints. Therefore, the limited and inconsistent results across
epidemiologic studies and the regional variation of air conditioning use has provided
inadequate evidence to determine whether a lower prevalence of air conditioning use
leads to a potential increased risk of O3-related health effects.
8.5 Summary
In this section, epidemiologic, controlled human exposure, and toxicological studies
have been evaluated and indicate that various factors may lead to increased risk of
O3-related health effects (Table 8-6).
The populations and lifestages identified in this section that have "adequate"
evidence for increased O3-related health effects are individuals with certain
genotypes, individuals with asthma, younger and older age groups, individuals with
reduced intake of certain nutrients, and outdoor workers, based on consistency in
findings across studies and evidence of coherence in results from different scientific
disciplines. Multiple genetic variants have been observed in epidemiologic and
controlled human exposure studies to affect the risk of O3-related respiratory
outcomes and support is provided by toxicological studies of genetic factors. Asthma
as a factor affecting risk was supported by controlled human exposure and
toxicological studies, as well as some evidence from epidemiologic studies.
Generally, studies of age groups reported positive associations for respiratory
hospital admissions and ED visits among children. Biological plausibility for this
increased risk is supported by toxicological and controlled human exposure research.
Also, children have higher exposure and dose due to increased time spent outdoors
and ventilation rate. Most studies comparing age groups reported greater effects of
short-term O3 exposure on mortality among older adults, although studies of other
health outcomes had inconsistent findings regarding whether older adults were at
increased risk. Older adults may also withstand greater O3 exposure and not seek
relief as quickly as younger adults. Multiple epidemiologic, controlled human
exposure, and toxicological studies reported that reduced Vitamins E and C intake
are associated with risk of O3-related health effects. Previous studies have shown that
increased exposure to O3 due to outdoor work leads to an increased risk of O3-related
health effects and it is clear that outdoor workers have higher exposures, and possibly
greater internal doses, of O3, which may lead to increased risk of O3-related health
effects.
8-35
-------
Table 8-6 Summary of evidence for potential increased risk of O3-related
health effects.
Evidence Classification Potential At Risk Factor
Adequate evidence Genetic factors (Section 8.1)
Asthma (Section 8.2.2)
Children (Section 8.3.1.1)
Older adults (Section 8.3.1.2)
Diet (Section 8.4.1)
Outdoor workers (Section 8.4.4)
Suggestive evidence Sex (Section 8.3.2)
SES (Section 8.3.3)
Obesity (Section 8.4.2)
Inadequate evidence Influenza/Infection (Section 8.2.1)
COPD (Section 8.2.3)
CVD (Section 8.2.4)
Diabetes (Section 8.2.5)
Hyperthyroidism (Section 8.2.6)
Race/ethnicity (Section 8.3.4)
Smoking (Section 8.4.3)
Air conditioning use (Section 8.4.5)
Evidence of no effect
In some cases, it is difficult to determine a factor that results in potentially increased
risk of effects. For example, previous assessments have included controlled human
exposure studies in which some healthy individuals demonstrate greater O3-related
health effects compared to other healthy individuals. Intersubject variability has been
observed for lung function decrements, symptomatic responses, pulmonary
inflammation, AHR, and altered epithelial permeability in healthy adults exposed to
O3 COueetal..2011: Holz et al. 2005: McDonnell. 1996). These responses to O3
exposure in healthy individuals tend to be reproducible within a given individual
over a period of several months indicating differences in the intrinsic responsiveness
(Holz et al.. 2005: Hazuchaet al.. 2003: Holzetal.. 1999: McDonnell et al.. 1985c).
In addition, there is the possibility of attenuation. In controlled human exposure and
toxicological studies, pre-exposure to O3 was observed to lead to a dampening of
some responses following subsequent exposure to O3 (for more details see
Sections 5.4.2.5 and 6.2.1.1).
Limitations include the challenge of evaluating effect measure modification in
epidemiologic studies with widespread populations with variation in numerous
factors. For a number of the factors described below, there are few available studies.
Also, some factors are inconsistent across studies, both in regards to the
categorization of the variable and its measurement in the studies. Many toxicological
and controlled human exposure studies are the only ones that have examined certain
factors and therefore have not been replicated. In considering epidemiologic studies
conducted in other countries, it is possible that those populations may differ in SES
8-36
-------
or other demographic indicators, thus limiting generalizability to a U.S. population.
Additionally, many epidemiologic studies that stratify by factors of interest have
small sample sizes, which can decrease precision of effect estimates and make
drawing conclusions difficult.
These challenges and limitations in evaluating the factors that can increase risk for
experiencing O3-related health effects may contribute to conclusions that evidence
for some factors, such as sex, SES, and obesity provided "suggestive" evidence of
potentially increased risk. In addition, for a number of factors listed in Table 8-6 the
evidence was inadequate to draw conclusions about potential increase in risk of
effects. Overall, the factors most strongly supported as contributing to increased risk
of O3-related effects among various populations and lifestages were related to
genetic factors, asthma, age group (children and older adults), dietary factors, and
working outdoors.
8-37
-------
References
Alexeeff. SE: Litonjua. AA; Suh. H; Sparrow. D; Vokonas. PS: Schwartz. J. (2007). Ozone exposure and lung
function: Effect modified by obesity and airways hyperresponsiveness in the VA Normative Aging Study.
Chest 132: 1890-1897. http://dx.doi.org/10.1378/chest.07-1126
Alexeeff. SE: Litonjua. AA; Wright. RO: Baccarelli. A; Suh. H; Sparrow. D; Vokonas. PS: Schwartz. J. (2008).
Ozone exposure, antioxidant genes, and lung function in an elderly cohort: VA Normative Aging Study.
Occup Environ Med 65: 736-742. http://dx.doi.org/10.1136/oem.2007.035253
Alexis. N; Urch. B; Tarlo. S: Corey. P: Pengelly. D: O'Byrne. P: Silverman. F. (2000). Cyclooxygenase
metabolites play a different role in ozone-induced pulmonary function decline in asthmatics compared to
normals. Inhal Toxicol 12: 1205-1224.
Alexis. NE; Zhou. H; Lav. JC: Harris. B; Hernandez. ML; Lu. TS: Bromberg. PA; Diaz-Sanchez. D; Devlin.
RB: Kleeberger. SR; Peden. DB. (2009). The glutathione-S-transferase Mu 1 null genotype modulates
ozone-induced airway inflammation in human subjects. J Allergy Clin Immunol 124: 1222-1228.
http://dx.doi.0rg/10.1016/i.iaci.2009.07.036
Arbex. AM; de Souza Conceicao. GM; Perez Cendon. S: Arbex. FF; Lopes. AC: Providello Moyses. E;
Santiago. SL; Saldiva. PHN; Pereira. LAA; Ferreira Braga. AL. (2009). Urban air pollution and COPD-
related emergency room visits. J Epidemiol Community Health 966: 777-783.
http://dx.doi.org/10.1136/iech.2008.078360
Auten. RL; Potts. EN; Mason. SN; Fischer. B; Huang. Y; Foster. WM. (2009). Maternal exposure to particulate
matter increases postnatal ozone-induced airway hyperreactivity in juvenile mice. Am J Respir Crit Care
Med 180: 1218-1226. http://dx.doi.org/10.1164/rccm.200901-0116OC
Avissar. NE; Reed. CK; Cox. C: Frampton. MW: Finkelstein. JN. (2000). Ozone, but not nitrogen dioxide,
exposure decreases glutathione peroxidases in epithelial lining fluid of human lung. Am J Respir Crit Care
Med 162: 1342-1347.
Baccarelli. A; Zanobetti. A; Martinelli. I; Grillo. P: Hou. L; Lanzani. G: Mannucci. PM; Bertazzi. PA;
Schwartz. J. (2007). Air pollution, smoking, and plasma homocysteine. Environ Health Perspect 115: 176-
181. http://dx.doi.org/10.1289/ehp.9517
Backus. GS: Howden. R; Fostel. J: Bauer. AK; Cho. HY; Marzec. J: Peden. DB; Kleeberger. SR. (2010).
Protective role of interleukin-10 in ozone-induced pulmonary inflammation. Environ Health Perspect 118:
1721-1727. http://dx.doi.org/10.1289/ehp.1002182
Barraza-Villarreal. A; Sunyer. J: Hernandez-Cadena. L; Escamilla-Nunez. MC: Sienra-Monge. JJ: Ramirez-
Aguilar. M; Cortez-Lugo. M; Holguin. F: Diaz-Sanchez. D; Olin. AC: Romieu. I. (2008). Air pollution,
airway inflammation, and lung function in a cohort study of Mexico City schoolchildren. Environ Health
Perspect 116: 832-838. http://dx.doi.org/10.1289/ehp.10926
Basha. MA; Gross. KB; Gwizdala. CJ: Haidar. AH; Popovich. J. Jr. (1994). Bronchoalveolar lavage
neutrophilia in asthmatic and healthy volunteers after controlled exposure to ozone and filtered purified air.
Chest 106: 1757-1765.
Bauer. AK; Rondini. EA; Hummel. KA; Degraff. LM; Walker. C: Jedlicka. AE; Kleeberger. SR. (2011).
Identification of candidate genes downstream of TLR4 signaling after ozone exposure in mice: A role for
heat shock protein 70. Environ Health Perspect 119: 1091-1097. http://dx.doi.org/10.1289/ehp.1003326
Bell. ML; Dominici. F. (2008). Effect modification by community characteristics on the short-term effects of
ozone exposure and mortality in 98 US communities. Am J Epidemiol 167: 986-997.
http://dx.doi.org/10.1093/aie/kwm396
8-38
-------
Bennett. WD; Hazucha. MJ; Folinsbee. LJ; Bromberg. PA; Kissling. GE; London. SJ. (2007). Acute
pulmonary function response to ozone in young adults as a function of body mass index. Inhal Toxicol 19:
1147-1154. http://dx.doi.org/10.1080/08958370701665475
Bergamaschi. E; De Palma. G: Mozzoni. P; Vanni. S: Vettori. MV; Broeckaert. F; Bernard. A; Mutti. A.
(2001). Polymorphism of quinone-metabolizing enzymes and susceptibility to ozone-induced acute effects.
Am J Respir Crit Care Med 163: 1426-1431.
Berhane. K; Zhang. Y; Linn. WS: Rappaport. EB; Bastain. TM; Salam. MT; Islam. T; Lurmann. F; Gilliland.
FD. (2011). The effect of ambient air pollution on exhaled nitric oxide in the Children's Health Study. Eur
Respir J 37: 1029-1036. http://dx.doi.org/10.1183/09031936.00081410
Bils. RF. (1970). Ultrastructural alterations of alveolar tissue of mice: III. Ozone. Arch Environ Health 20:
468-480.
Blomberg. A; Mudwav. IS: Nordenhall. C: Hedenstrom. H; Kelly. FJ; Frew. AJ; Holgate. ST; Sandstrom. T.
(1999). Ozone-induced lung function decrements do not correlate with early airway inflammatory or
antioxidant responses. Eur Respir J 13: 1418-1428.
Bloom. B; Cohen. RA; Freeman. G. (2008). Summary health statistics for U.S. children: National Health
Interview Survey, 2008 (pp. 90). Washington, DC: National Center for Health Statistics.
Bosson. J; Stenfors. N; Bucht. A; Helledav. R; Pourazar. J: Holgate. ST; Kelly. FJ: Sandstrom. T; Wilson. S:
Frew. AJ; Blomberg. A. (2003). Ozone-induced bronchial epithelial cytokine expression differs between
healthy and asthmatic subjects. Clin Exp Allergy 33: 777-782.
Broeckaert. F; Clippe. A; Wattiez. R; Falmagne. P; Bernard. A. (2003). Lung hyperpermeability, Clara-cell
secretory potein (CC16), and susceptibility to ozone of five inbred strains of mice. Inhal Toxicol 15: 1209-
1230.
Buadong. D; Jinsart. W: Funatagawa. I; Karita. K; Yano. E. (2009). Association between PM10 and O3 levels
and hospital visits for cardiovascular diseases in Bangkok, Thailand. J Epidemiol 19: 182-188.
http://dx.doi.org/10.2188/iea.JE20080047
Burra. TA; Moineddin. R; Agha. MM; Glazier. RH. (2009). Social disadvantage, air pollution, and asthma
physician visits in Toronto, Canada. Environ Res 109: 567-574.
http://dx.doi.0rg/10.1016/i.envres.2009.03.004
Bush. ML; Asplund. PT; Miles. KA; Ben-Jebria. A; Ultman. JS. (1996). Longitudinal distribution of O3
absorption in the lung: gender differences and intersubject variability. J Appl Physiol 81: 1651-1657.
Cakmak. S: Dales. RE; Angelica Rubio. M; Blanco Vidal. C. (2011). The risk of dying on days of higher air
pollution among the socially disadvantaged elderly. Environ Res 111: 388-393.
http://dx.doi.0rg/10.1016/i.envres.2011.01.003
Cakmak. S: Dales. RE; Judek. S. (2006a). Do gender, education, and income modify the effect of air pollution
gases on cardiac disease? J Occup Environ Med 48: 89-94.
http://dx.doi.org/10.1097/01.iom.0000184878.11956.4b
Cakmak. S: Dales. RE; Judek. S. (2006b). Respiratory health effects of air pollution gases: Modification by
education and income. Arch Environ Occup Health 61: 5-10. http://dx.doi.Org/10.3200/AEOH.61.l.5-10
Cakmak. S: Dales. RE; Vidal. CB. (2007). Air pollution and mortality in Chile: Susceptibility among the
elderly. Environ Health Perspect 115: 524-527.
Calderon Guzman. D; Barragan Mejia. G: Hernandez Garcia. E; Juarez Olguin. H. (2006). Effect of nutritional
status and ozone exposure on some biomarkers of oxidative stress in rat brain regions. Nutr Cancer 55:
195-200. http://dx.doi.org/10.1207/sl5327914nc5502 11
Carbaial-Arroyo. L; Miranda-Soberanis. V: Medina-Ramon. M; Rojas-Bracho. L; Tzintzun. G: Solis-
Gutierrez. P; Mendez-Ramirez. I; Hurtado-Diaz. M; Schwartz. J: Romieu. I. (2011). Effect of PM10 and
O3 on infant mortality among residents in the Mexico City Metropolitan Area: A case-crossover analysis,
19972005. J Epidemiol Community Health 65: 715-721. http://dx.doi.org/10.1136/iech.2009.101212
8-39
-------
Carey. SA; Minard. KR; Trease. LL; Wagner. JG: Garcia. GJ: Ballinger. CA; Kimbell. JS: Plopper. CG: Corlev.
RA; Postlethwait. EM; Harkema. JR. (2007). Three-dimensional mapping of ozone-induced injury in the
nasal airways of monkeys using magnetic resonance imaging and morphometric techniques. Toxicol Pathol
35: 27-40. http://dx.doi.org/10.1080/01926230601072343
Chhabra. SK; Yasir. A; Chaudhry. K; Shah. B. (2010). Effect of ozone on response to ovalbumin & its
modulation by vitamins C & E in sensitized guinea pigs. Indian J Med Res 132: 87-93.
Chiu. HF; Yang. CY. (2009). Air pollution and emergency room visits for arrhythmias: Are there potentially
sensitive groups? J Toxicol Environ Health A 72: 817-823. http://dx.doi.org/10.1080/15287390902800405
Cho. HY; Morgan. PL; Bauer. AK; Kleeberger. SR. (2007). Signal transduction pathways of tumor necrosis
factor—mediated lung injury induced by ozone in mice. Am J Respir Crit Care Med 175: 829-839.
http://dx.doi.org/10.1164/rccm.200509-1527OC
Cho. HY; Zhang. LY; Kleeberger. SR. (2001). Ozone-induced lung inflammation and hyperreactivity are
mediated via tumor necrosis factor-alpha receptors. Am J Physiol 280: L537-L546.
Chuang. GC: Yang. Z; Westbrook. DG: Pompilius. M; Ballinger. CA; White. RC: Krzywanski. DM;
Postlethwait. EM; Ballinger. SW. (2009). Pulmonary ozone exposure induces vascular dysfunction,
mitochondria! damage, and atherogenesis. Am J Physiol Lung Cell Mol Physiol 297: L209-L216.
http://dx.doi.org/10.1152/aiplung.00102.2009
Coffin. PL; Blommer. EJ; Gardner. DE; Holzman. R. (1967). Effect of air pollution on alteration of
susceptibility to pulmonary infection (pp. 71-80). (AMRL-TR-67-200). Cincinnati, OH: Aerospace Medical
Research Laboratories.
Coffin. PL; Gardner. DE. (1972). Interaction of biological agents and chemical air pollutants. Ann Occup Hyg
15: 219-234.
Dahl. M; Bauer. AK; Arredouani. M; Soininen. R; Trvggvason. K; Kleeberger. SR: Kobzik. L. (2007).
Protection against inhaled oxidants through scavenging of oxidized lipids by macrophage receptors
MARCO and SR-AI/II. J Clin Invest 117: 757-764. http://dx.doi.org/10.1172/JCI29968
Darrow. LA; Klein. M; Strickland. MJ; Mulholland. JA; Tolbert. PE. (20 lib). Ambient air pollution and birth
weight in full-term infants in Atlanta, 1994-2004. Environ Health Perspect 119: 731-737.
http://dx.doi.org/10.1289/ehp.1002785
Escamilla-Nunez. MC: Barraza-Villarreal. A; Hernandez-Cadena. L; Moreno-Macias. H; Ramirez-Aguilar. M;
Sienra-Monge. JJ: Cortez-Lugo. M; Texcalac. JL; del Rio-Navarro. B; Romieu. I. (2008). Traffic-related air
pollution and respiratory symptoms among asthmatic children, resident in Mexico City: The EVA cohort
study. Respir Res 9: 74. http://dx.doi.org/10.1186/1465-9921-9-74
Fakhrzadeh. L; Laskin. JD; Laskin. PL. (2002). Deficiency in inducible nitric oxide synthase protects mice
from ozone-induced lung inflammation and tissue injury. Am J Respir Cell Mol Biol 26: 413-419.
Fanucchi. MV; Plopper. CG: Evans. MJ: Hyde. DM; Van Winkle. LS: Gershwin. LJ; Schelegle. ES. (2006).
Cyclic exposure to ozone alters distal airway development in infant rhesus monkeys. Am J Physiol Lung
Cell Mol Physiol 291: L644-L650. http://dx.doi.org/10.1152/aiplung.00027.2006
Fortino. V; Maioli. E; Torricelli. C: Davis. P; Valacchi. G. (2007). Cutaneous MMPs are differently modulated
by environmental stressors in old and young mice. Toxicol Lett 173: 73-79.
http://dx.doi.0rg/10.1016/i.toxlet.2007.06.004
Fox. SD; Adams. WC: Brookes. KA; Laslev. BL. (1993). Enhanced response to ozone exposure during the
follicular phase of the menstrual cycle. Environ Health Perspect 101: 242-244.
Frampton. MW: Morrow. PE; Torres. A; Cox. C: Voter. KZ; Utell. MJ: Gibb. FR; Speers. DM. (1997a). Ozone
responsiveness in smokers and nonsmokers. Am J Respir Crit Care Med 155: 116-121.
Funabashi. H; Shima. M; Kuwaki. T; Hiroshima. K; Kuriyama. T. (2004). Effects of repeated ozone exposure
on pulmonary function and bronchial responsiveness in mice sensitized with ovalbumin. Toxicology 204:
75-83. http://dx.doi.0rg/10.1016/i.tox.2004.06.047
8-40
-------
Garantziotis. S: Li. Z; Potts. EN; Kimata. K; Zhuo. L; Morgan. PL; Savani. RC: Noble. PW: Foster. WM;
Schwartz. DA; Hollingsworth. JW. (2009). Hyaluronan mediates ozone-induced airway
hyperresponsiveness in mice. J Biol Chem 284: 11309-11317. http://dx.doi.org/10.1074/ibc.M802400200
Garantziotis. S: Li. Z; Potts. EN; Lindsev. JY; Stober. VP; Polosukhin. W; Blackwell. TS: Schwartz. DA;
Foster. WM; Hollingsworth. JW. (2010). TLR4 is necessary for hyaluronan-mediated airway
hyperresponsiveness after ozone inhalation. Am J Respir Crit Care Med 181: 666-675.
http://dx.doi.org/10.1164/rccm.200903-0381OC
Goldberg. MS: Burnett. RT; Yale. JF; Valois. MF; Brook. JR. (2006). Associations between ambient air
pollution and daily mortality among persons with diabetes and cardiovascular disease. Environ Res 100:
255-267.
Halonen. JI; Lanki. T; Tiittanen. P; Niemi. JV: Loh. M; Pekkanen. J. (2009). Ozone and cause-specific
cardiorespiratory morbidity and mortality. J Epidemiol Community Health 64: 814-820.
http://dx.doi.org/10.1136/iech.2009.087106
Hamade. AK; Misra. V: Rabold. R; Tankerslev. CG. (2010). Age-related changes in cardiac and respiratory
adaptation to acute ozone and carbon black exposures: Interstrain variation in mice. Inhal Toxicol 22: 84-
94. http://dx.doi.org/10.3109/08958378.2010.503974
Hamade. AK; Rabold. R; Tankerslev. CG. (2008). Adverse cardiovascular effects with acute paniculate matter
and ozone exposures: Interstrain variation in mice. Environ Health Perspect 116: 1033-1039.
http://dx.doi.org/10.1289/ehp.10689
Hamade. AK: Tankerslev. CG. (2009). Interstrain variation in cardiac and respiratory adaptation to repeated
ozone and paniculate matter exposures. Am J Physiol Regul Integr Comp Physiol 296: R1202-R1215.
http://dx.doi.org/10.1152/ajpregu.90808.2008
Hansen. CA; Barnett. AG: Pritchard. G. (2008). The effect of ambient air pollution during early pregnancy on
fetal ultrasonic measurements during mid-pregnancy. Environ Health Perspect 116: 362-369.
http://dx.doi.org/10.1289/ehp.10720
Harkema. JR; Plopper. CG: Hyde. DM; St George. JA; Dungworth. PL. (1987a). Effects of an ambient level of
ozone on primate nasal epithelial mucosubstances: quantitative histochemistry. Am J Pathol 127: 90-96.
Hazucha. MJ; Folinsbee. LJ; Bromberg. PA. (2003). Distribution and reproducibility of spirometric response to
ozone by gender and age. J Appl Physiol 95: 1917-1925.
Henrotin. JB; Besancenot. JP; Bejot. Y; Giroud. M. (2007). Short-term effects of ozone air pollution on
ischaemic stroke occurrence: A case-crossover analysis from a 10-year population-based study in Dijon,
France. Occup Environ Med 64: 439-445. http://dx.doi.org/10.1136/oem.2006.029306
Henrotin. JB; Zeller. M; Lorgis. L; Cottin. Y; Giroud. M; Bejot. Y. (2010). Evidence of the role of short-term
exposure to ozone on ischaemic cerebral and cardiac events: The Dijon Vascular Project (DIVA). Heart 96:
1990-1996. http://dx.doi.org/10.1136/hrt.2010.200337
Hernandez-Cadena. L; Holguin. F; Barraza-Villarreal. A; Del Rio-Navarro. BE; Sienra-Monge. JJ: Romieu. I.
(2009). Increased levels of outdoor air pollutants are associated with reduced bronchodilation in children
with asthma. Chest 136: 1529-1536. http://dx.doi.org/10.1378/chest.08-1463
Hernandez. ML; Lav. JC: Harris. B; Esther. CR; Brickev. WJ: Bromberg. PA; Diaz-Sanchez. D; Devlin. RB;
Kleeberger. SR; Alexis. NE; Peden. DB. (2010). Atopic asthmatic subjects but not atopic subjects without
asthma have enhanced inflammatory response to ozone. J Allergy Clin Immunol 126: 537-544.
http://dx.doi.0rg/10.1016/i.jaci.2010.06.043
Hollingsworth. JW: Cook. DN; Brass. DM; Walker. JKL; Morgan. PL; Foster. WM; Schwartz. DA. (2004).
The role of Toll-like receptor 4 in environmental airway injury in mice. Am J Respir Crit Care Med 170:
126-132. http://dx.doi.org/10.1164/rccm.200311-1499OC
Hollingsworth. JW: Free. ME; Li. Z; Andrews. LN; Nakano. H; Cook. DN. (2010). Ozone activates pulmonary
dendritic cells and promotes allergic sensitization through a Toll-like receptor 4-dependent mechanism
[Letter]. J Allergy Clin Immunol 125: 1167-1170. http://dx.doi.Org/10.1016/i.iaci.2010.03.001
8-41
-------
Holz. O; Jorres. RA; Timm. P; Mucke. M; Richter. K; Koschyk. S: Magnussen. H. (1999). Ozone-induced
airway inflammatory changes differ between individuals and are reproducible. Am J Respir Crit Care Med
159: 776-784.
Holz. O; Tal-Singer. R; Kanniess. F; Simpson. KJ; Gibson. A; Vessev. RSJ: Janicki. S: Magnussen. H; Jorres.
RA; Richter. K. (2005). Validation of the human ozone challenge model as a tool for assessing anti-
inflammatory drugs in early development. J Clin Pharmacol 45: 498-503.
Horstman. DH; Ball. BA; Brown. J; Gerritv. T; Folinsbee. LJ. (1995). Comparison of pulmonary responses of
asthmatic and nonasthmatic subjects performing light exercise while exposed to a low level of ozone.
Toxicol Ind Health 11: 369-385.
Houslev. DG: Eccles. R; Richards. RJ. (1996). Gender difference in the concentration of the antioxidant uric
acid in human nasal lavage. Acta Otolaryngol 116: 751-754.
Hu. SC: Ben-Jebria. A; Ultman. JS. (1994). Longitudinal distribution of ozone absorption in the lung: Effects
of respiratory flow. J Appl Physiol 77: 574-583.
Huffman. LJ: Beighlev. CM; Frazer. DG: Mckinnev. WG: Porter. DW. (2006). Increased susceptibility of the
lungs of hyperthyroid rats to oxidant injury: Specificity of effects. Toxicology 225: 119-127.
http://dx.doi.0rg/10.1016/i.tox.2006.05.008
Jerrett. M; Burnett. RT; Pope. CA. Ill; Ito. K; Thurston. G: Krewski. D; Shi. Y; Calle. E; Thun. M. (2009).
Long-term ozone exposure and mortality. N Engl J Med 360: 1085-1095.
http://dx.doi.org/10.1056/NEJMoa0803894
Joad. JP: Kott. KS: Brie. JM; Peake. JL; Plopper. CG: Schelegle. ES: Gershwin. LJ: Pinkerton. KE. (2006).
Structural and functional localization of airway effects from episodic exposure of infant monkeys to
allergen and/or ozone. Toxicol Appl Pharmacol 214: 237-243. http://dx.doi.Org/10.1016/i.taap.2005.12.012
Johnston. RA; Mizgerd. JP: Shore. SA. (2005a). CXCR2 is essential for maximal neutrophil recruitment and
methacholine responsiveness after ozone exposure. Am J Physiol Lung Cell Mol Physiol 288: L61-L67.
http://dx.doi.org/10.1152/aiplung.00101.2004 00101.2004
Johnston. RA: Schwartzman. IN; Flynt. L; Shore. SA. (2005b). Role of interleukin-6 in murine airway
responses to ozone. Am J Physiol Lung Cell Mol Physiol 288: L390-L397.
http://dx.doi.org/10.1152/aiplung.00007.2004
Johnston. RA; Theman. TA; Lu. FL; Terry. RD; Williams. ES: Shore. SA. (2008). Diet-induced obesity causes
innate airway hyperresponsiveness to methacholine and enhances ozone-induced pulmonary inflammation.
J Appl Psychol 104: 1727-1735. http://dx.doi.org/10.1152/iapplphysiol.00075.2008
Jorres. R: Nowak. D; Magnussen. H; Speckin. P; Koschyk. S. (1996). The effect of ozone exposure on allergen
responsiveness in subjects with asthma or rhinitis. Am J Respir Crit Care Med 153: 56-64.
Kan. H; London. SJ: Chen. G: Zhang. Y; Song. G: Zhao. N; Jiang. L; Chen. B. (2008). Season, sex, age, and
education as modifiers of the effects of outdoor air pollution on daily mortality in Shanghai, China:
The Public Health and Air Pollution in Asia (PAPA) Study. Environ Health Perspect 116: 1183-1188.
http://dx.doi.org/10.1289/ehp.10851
Kari. F; Hatch. G: Slade. R: Crissman. K; Simeonova. PP; Luster. M. (1997). Dietary restriction mitigates
ozone-induced lung inflammation in rats: A role for endogenous antioxidants. Am J Respir Cell Mol Biol
17: 740-747.
Katsouvanni. K; Samet. JM; Anderson. HR; Atkinson. R; Le Tertre. A; Medina. S: Samoli. E; Touloumi. G:
Burnett. RT; Krewski. D; Ramsay. T; Dominici. F; Peng. RD; Schwartz. J: Zanobetti. A. (2009). Air
pollution and health: A European and North American approach (APHENA). (Research Report 142).
Boston, MA: Health Effects Institute. http://pubs.healtheffects.org/view.php?id=327
Kehrl. HR; Peden. DB; Ball. BA; Folinsbee. LJ: Horstman. DH. (1999). Increased specific airway reactivity of
persons with mild allergic asthma after 7.6 hours of exposure to 0.16 ppm ozone. J Allergy Clin Immunol
104: 1198-1204.
8-42
-------
Kenyon. NJ; Van Per Vliet A; Schock. BC: Okamoto. T; McGrew. GM; Last. JA. (2002). Susceptibility to
ozone-induced acute lung injury in iNOS-deficient mice. Am J Physiol 282: L540-L545.
Khatri. SB; Holguin. FC: Ryan. PB; Manning. D; Erzurum. SC: Teague. WG. (2009). Association of ambient
ozone exposure with airway inflammation and allergy in adults with asthma. J Asthma 46: 777-785.
http://dx.doi.org/10.1080/02770900902779284
Kim. CS: Alexis. NE; Rappold. AG: Kehrl. H; Hazucha. MJ; Lav. JC: Schmitt. MT; Case. M; Devlin. RB;
Peden. DB; Diaz-Sanchez. D. (2011). Lung function and inflammatory responses in healthy young adults
exposed to 0.06 ppm ozone for 6.6 hours. Am J Respir Crit Care Med 183: 1215-1221.
http://dx.doi.org/10.1164/rccm.201011-1813OC
Kleeberger. SR; Levitt. RC: Zhang. LY; Longphre. M; Harkema. J: Jedlicka. A; Eleff. SM; DiSilvestre. D:
Holroyd. KJ. (1997). Linkage analysis of susceptibility to ozone-induced lung inflammation in inbred mice.
Nat Genet 17: 475-478.
Kleeberger. SR: Reddv. S: Zhang. LY; Jedlicka. AE. (2000). Genetic susceptibility to ozone-induced lung
hyperpermeability: Role of toll-like receptor 4. Am J Respir Cell Mol Biol 22: 620-627.
Kleeberger. SR: Reddv. SP; Zhang. LY; Cho. HY; Jedlicka. AE. (2001). Toll-like receptor 4 mediates ozone-
induced murine lung hyperpermeability via inducible nitric oxide synthase. Am J Physiol 280: L326-L333.
Ko. FWS: Tarn. W: Wong. TW: Lai. CKW. (2007). Effects of air pollution on asthma hospitalization rates in
different age groups in Hong Kong. Clin Exp Allergy 37: 1312-1319. http://dx.doi.org/10.llll/U365-
2222.2007.02791.x
Kodavanti. UP; Costa. PL; Dreher. KL; Crissman. K; Hatch. GE. (1995). Ozone-induced tissue injury and
changes in antioxidant homeostasis in normal and ascorbate-deficient guinea pigs. Biochem Pharmacol 50:
243-251. http://dx.doi.org/10.1016/0006-2952(95)00122-G
Kooter. IM; Pennings. JL; Fokkens. PH; Leseman. PL; Boere. AJ; Gerlofs-Nijland. ME; Cassee. FR; Schalk.
JA; Orzechowski. TJ; Schaap. MM; Breit TM; Dormans. JA; van Oostrom. CT; de Vries. A; van Steeg. H.
(2007). Ozone induces clear cellular and molecular responses in the mouse lung independently of the
transcription-coupled repair status. JAppl Physiol 102: 1185-1192.
http://dx.doi.org/10.1152/iapplphvsiol.00796.2006
Kreit. JW: Gross. KB; Moore. TB; Lorenzen. TJ: D'Arcv. J: Eschenbacher. WL. (1989). Ozone-induced
changes in pulmonary function and bronchial responsiveness in asthmatics. JAppl Physiol 66: 217-222.
Kumarathasan. P; Blais. E; Goegan. P; Yagminas. A; Guenette. J: Adamson. IY; Crapo. JD; Mason. RJ;
Vincent. R. (2005). 90-day repeated inhalation exposure of surfactant Protein-C/tumor necrosis factor-
alpha, (SP-C/TNF-alpha) transgenic mice to air pollutants. Int J Toxicol 24: 59-67.
Lagorio. S: Forastiere. F; Pistelli. R; lavarone. I; Michelozzi. P; Fano. V; Marconi. A; Ziemacki. G: Ostro. BD.
(2006). Air pollution and lung function among susceptible adult subjects: A panel study. Environ Health 5:
11. http://dx.doi.org/10.1186/1476-069X-5-ll
Lee. IM; Tsai. SS: Ho. CK; Chiu. HF; Wu. TN; Yang. CY. (2008a). Air pollution and hospital admissions for
congestive heart failure: Are there potentially sensitive groups? Environ Res 108: 348-353.
http://dx.doi.0rg/10.1016/i.envres.2008.07.024
Lee. JT; Son. JY; Kim. H; Kim. SY. (2006). Effect of air pollution on asthma-related hospital admissions for
children by socioeconomic status associated with area of residence. Arch Environ Occup Health 61: 123-
130.
Lewis. TC: Robins. TG: Dvonch. JT: Keeler. GJ: Yip. FY: Mentz. GB: Lin. X: Parker. EA: Israel. BA:
Gonzalez. L; Hill. Y (2005). Air pollution-associated changes in lung function among asthmatic children in
Detroit. Environ Health Perspect 113: 1068-1075.
Liao. D: Heiss. G: Chinchilli. VM; Duan. Y; Folsom. AR; Lin. HM; Salomaa. V. (2005). Association of criteria
pollutants with plasma hemostatic/inflammatory markers: A population-based study. J Expo Sci Environ
Epidemiol 15: 319-328.
8-43
-------
Lim. Y; Phung. AD; Corbacho. AM; Aung. HH; Maioli. E; Reznick. AZ; Cross. CE; Davis. PA; Valacchi. G.
(2006). Modulation of cutaneous wound healing by ozone: Differences between young and aged mice.
Toxicol Lett 160: 127-134. http://dx.doi.Org/10.1016/j.toxlet.2005.06.013
Lin. M; Stieb. DM; Chen. Y. (2005). Coarse paniculate matter and hospitalization for respiratory infections in
children younger than 15 years in Toronto: A case-crossover analysis. Pediatrics 116: 235-240.
Lin. S: Liu. X; Le. LH; Hwang. SA. (2008b). Chronic exposure to ambient ozone and asthma hospital
admissions among children. Environ Health Perspect 116: 1725-1730. http://dx.doi.org/10.1289/ehp.11184
Liu. L; Poon. R; Chen. L; Frescura. AM; Montuschi. P; Ciabattoni. G: Wheeler. A; Dales. R. (2009a). Acute
effects of air pollution on pulmonary function, airway inflammation, and oxidative stress in asthmatic
children. Environ Health Perspect 117: 668-674. http://dx.doi.org/10.1289/ehpll813
Lopez. I; Sanchez. I; Bizarro. P; Acevedo. S: Ustarroz. M; Fortoul. T. (2008). Ultrastructural alterations during
embryonic rats' lung development caused by ozone. J Electron Microsc (Tokyo) 57: 19-23.
http://dx.doi.org/10.1093/imicro/dfm033
Mar. TF; Koenig. JQ. (2009). Relationship between visits to emergency departments for asthma and ozone
exposure in greater Seattle, Washington. Ann Allergy Asthma Immunol 103: 474-479.
http://dx.doi.org/10.1016/S1081-1206(10)60263-3
McDonnell. WF. (1996). Individual variability in human lung function responses to ozone exposure. Environ
Toxicol Pharmacol 2: 171-175.
McDonnell. WF. Ill; Horstman. DH; Abdul-Salaam. S: House. DE. (1985c). Reproducibility of individual
responses to ozone exposure. Am Rev Respir Dis 131: 36-40.
McDonnell. WF: Stewart. PW: Smith. MV. (2010). Prediction of ozone-induced lung function responses in
humans. Inhal Toxicol 22: 160-168. http://dx.doi.org/10.3109/08958370903089557
McDonnell. WF: Stewart. PW: Smith. MV: Pan. WK; Pan. J. (1999b). Ozone-induced respiratory symptoms:
Exposure-response models and association with lung function. Eur Respir J 14: 845-853.
Medina-Ramon. M; Schwartz. J. (2008). Who is more vulnerable to die from ozone air pollution?
Epidemiology 19: 672-679. http://dx.doi.org/10.1097/EDE.Ob013e3181773476
Medina-Ramon. M; Zanobetti. A; Schwartz. J. (2006). The effect of ozone and PM10 on hospital admissions
for pneumonia and chronic obstructive pulmonary disease: A national multicity study. Am J Epidemiol 163:
579-588. http://dx.doi.org/10.1093/aje/kwj078
Middleton. N; Yiallouros. P: Kleanthous. S: Kolokotroni. O; Schwartz. J: Dockerv. DW: Demokritou. P:
Koutrakis. P. (2008). A 10-year time-series analysis of respiratory and cardiovascular morbidity in Nicosia,
Cyprus: The effect of short-term changes in air pollution and dust storms. Environ Health 7: 39.
Miller. FJ; Illing. JW: Gardner. DE. (1978). Effect of urban ozone levels on laboratory-induced respiratory
infections. Toxicol Lett 2: 163-169.
Morello-Frosch. R; Jesdale. BM; Sadd. JL; Pastor. M. (2010). Ambient air pollution exposure and full-term
birth weight in California. Environ Health 9: 44. http://dx.doi.org/10.1186/1476-069X-9-44
Mudwav. IS: Blomberg. A; Frew. AJ; Holgate. ST; Sandstrom. T; Kelly. FJ. (1999a). Antioxidant consumption
and repletion kinetics in nasal lavage fluid following exposure of healthy human volunteers to ozone. Eur
Respir J 13: 1429-1438.
Mudwav. IS: Stenfors. N; Blomberg. A; Helledav. R: Punster. C: Marklund. SL; Frew. AJ: Sandstrom. T;
Kelly. FJ. (2001). Differences in basal airway antioxidant concentrations are not predictive of individual
responsiveness to ozone: A comparison of healthy and mild asthmatic subjects. Free Radic Biol Med 31:
962-974.
Neidell. M; Kinnev. PL. (2010). Estimates of the association between ozone and asthma hospitalizations that
account for behavioral responses to air quality information. Environ Sci Pol 13: 97-103.
http://dx.doi.0rg/10.1016/i.envsci.2009.12.006
8-44
-------
Nodelman. V; Ultman. JS. (1999). Longitudinal distribution of chlorine absorption in human airways:
A comparison to ozone absorption. J Appl Physiol 87: 2073-2080.
Ovarzun. M; Dussaubat. N; Gonzalez. S. (2005). Effect of 0.25 ppm ozone exposure on pulmonary damage
induced by bleomycin. Biol Res 38: 353-358.
Palli. D: Sera. F; Giovannelli. L; Masala. G: Grechi. D; Bendinelli. B; Caini. S: Dolara. P; Saieva. C. (2009).
Environmental ozone exposure and oxidative DNA damage in adult residents of Florence, Italy. Environ
Pollut 157: 1521-1525. http://dx.doi.Org/10.1016/i.envpol.2008.09.011
Paquette. NC: Zhang. LY; Ellis. WA; Scott. AL; Kleeberger. SR. (1996). Vitamin A deficiency enhances ozone-
induced lung injury. Am J Physiol 270: L475-L482.
Paulu. C: Smith. AE. (2008). Tracking associations between ambient ozone and asthma-related emergency
department visits using case-crossover analysis. J Public Health Manag Pract 14: 581-591.
Peden. DB; Boehlecke. B; Horstman. D; Devlin. R. (1997). Prolonged acute exposure to 0.16 ppm ozone
induces eosinophilic airway inflammation in asthmatic subjects with allergies. J Allergy Clin Immunol 100:
802-808.
Peden. DB; Setzer. RW. Jr; Devlin. RB. (1995). Ozone exposure has both a priming effect on allergen-induced
responses and an intrinsic inflammatory action in the nasal airways of perennially allergic asthmatics. Am J
Respir Crit Care Med 151: 1336-1345.
Peel. JL; Metzger. KB; Klein. M; Flanders. WD; Mulholland. JA; Tolbert. PE. (2007). Ambient air pollution
and cardiovascular emergency department visits in potentially sensitive groups. Am J Epidemiol 165: 625-
633. http://dx.doi.org/10.1093/aie/kwk051
Pleis. JR: Lucas. JW: Ward. BW. (2009). Summary health statistics for U.S. adults: National Health Interview
Survey, 2008. (DHHS 2010-1570). Hyattsville, MD: National Center for Health Statistics.
http://www.cdc.gov/nchs/data/series/sr 10/srlO 242.pdf
Plopper. CG: Smiley-Jewell. SM; Miller. LA; Fanucchi. MV; Evans. MJ; Buckpitt. AR; Avdalovic. M;
Gershwin. LJ; Joad. JP; Kaiekar. R; Larson. S: Pinkerton. KB; Van Winkle. LS: Schelegle. ES: Pieczarka.
EM; Wu. R; Hyde. DM. (2007). Asthma/allergic airways disease: Does postnatal exposure to
environmental toxicants promote airway pathobiology? Toxicol Pathol 35: 97-110.
http://dx.doi.org/10.1080/01926230601132030
Qian. Z; Lin. HM; Chinchilli. VM; Lehman. EB; Duan. Y; Craig. TJ; Wilson. WE; Liao. D: Lazarus. SC:
Bascom. R. (2009). Interaction of ambient air pollution with asthma medication on exhaled nitric oxide
among asthmatics. Arch Environ Occup Health 64: 168-176.
http://dx.doi.org/10.1080/19338240903240616
Que. LG: Stiles. JV: Sundv. JS: Foster. WM. (2011). Pulmonary function, bronchial reactivity, and epithelial
permeability are response phenotypes to ozone and develop differentially in healthy humans. J Appl
Physiol 111: 679-687. http://dx.doi.org/10.1152/iapplphysiol.00337.2011
Rivas-Arancibia. S: Dorado-Martinez. C: Borgonio-Perez. G: Hiriart-Urdanivia. M; Verdugo-Diaz. L; Duran-
Vazquez. A; Colin-Baranque. L; Avila-Costa. MR. (2000). Effects of taurine on ozone-induced memory
deficits and lipid peroxidation levels in brains of young, mature, and old rats. Environ Res 82: 7-17.
http://dx.doi.org/10.1006/enrs.1999.3996
Romieu. I; Barraza-Villarreal. A; Escamilla-Nunez. C: Texcalac-Sangrador. JL; Hernandez-Cadena. L; Diaz-
Sanchez. D; De Batlle. J: Del Rio-Navarro. BE. (2009). Dietary intake, lung function and airway
inflammation in Mexico City school children exposed to air pollutants. Respir Res 10: 122.
Romieu. I; Ramirez-Aguilar. M; Moreno-Macias. H; Barraza-Villarreal. A; Miller. P; Hernandez-Cadena. L;
Carbaial-Arroyo. LA; Hernandez-Avila. M. (2004a). Infant mortality and air pollution: Modifying effect by
social class. J Occup Environ Hyg 46: 1210-1216.
8-45
-------
Romieu. I; Ramirez-Aguilar. M; Sienra-Monge. JJ: Moreno-Macias. H; Del Rio-Navarro. BE; David. G:
Marzec. J: Hernandez-Avila. M; London. S. (2006). GSTM1 and GSTP1 and respiratory health in
asthmatic children exposed to ozone. Eur Respir J 28: 953-959.
http://dx.doi.org/10.1183/09031936.06.00114905
Romieu. I; Sienra-Monge. JJ: Ramirez-Aguilar. M; Moreno-Macias. H; Reyes-Ruiz. NI; Estela del Rio-
Navarro. B; Hernandez-Avila. M; London. SJ. (2004b). Genetic polymorphism of GSTM1 and antioxidant
supplementation influence lung function in relation to ozone exposure in asthmatic children in Mexico City.
Thorax 59: 8-10.
Ruidavets. JB; Cassadou. S: Cournot. M; Bataille. V: Mevbeck. M; Ferrieres. J. (2005a). Increased resting
heart rate with pollutants in a population based study. J Epidemiol Community Health 59: 685-693.
Sacks. JD: Stanek. LW: Luben. TJ; Johns. DO: Buckley. BJ; Brown. JS: Ross. M. (2011). Particulate-matter
induced health effects: Who is susceptible? Environ Health Perspect 119: 446-454.
http://dx.doi.org/10.1289/ehp.1002255
Samet. JM; Hatch. GE; Horstman. D; Steck-Scott. S: Arab. L; Bromberg. PA; Levine. M; McDonnell. WF;
Devlin. RB. (2001). Effect of antioxidant supplementation on ozone-induced lung injury in human subjects.
Am J Respir Crit Care Med 164: 819-825.
Sarangapani. R; Gentry. PR; Covington. TR; Teeguarden. JG: Clewell HJ. III. (2003). Evaluation of the
potential impact of age- and gender-specific lung morphology and ventilation rate on the dosimetry of
vapors. Inhal Toxicol 15: 987-1016. http://dx.doi.org/10.1080/08958370390226350
Sawyer. K; Brown. J: HazuchaM; Bennett. WD. (2007). The effect of exercise on nasal uptake of ozone in
healthy human adults. J Appl Physiol 102: 1380-1386. http://dx.doi.org/10.1152/iapplphysiol.00269.2006
Scannell. C: Chen. L; Aris. RM; Tager. I; Christian. D; Ferrando. R; Welch. B; Kelly. T; Balmes. JR. (1996).
Greater ozone-induced inflammatory responses in subjects with asthma. Am J Respir Crit Care Med 154:
24-29.
Schelegle. ES: Miller. LA; Gershwin. LJ; Fanucchi. MV; Van Winkle. LS: Gerriets. JE; Walbv. WF; Mitchell.
V; Tarkington. BK; Wong. VJ; Baker. GL; Pantle. LM; Joad. JP; Pinkerton. KE; Wu. R: Evans. MJ; Hyde.
DM; Plopper. CG. (2003). Repeated episodes of ozone inhalation amplifies the effects of allergen
sensitization and inhalation on airway immune and structural development in Rhesus monkeys. Toxicol
Appl Pharmacol 191: 74-85.
Seal. E. Jr; McDonnell. WF: House. DE. (1996). Effects of age, socioeconomic status, and menstrual cycle on
pulmonary response to ozone. Arch Environ Occup Health 51: 132-137.
Seal. E. Jr; McDonnell. WF: House. DE; Salaam. SA; Dewitt. PJ: Butler. SO: Green. J: Raggio. L. (1993).
The pulmonary response of white and black adults to six concentrations of ozone. Am J Respir Crit Care
Med 147: 804-810.
Servais. S: Boussouar. A; Molnar. A; Douki. T; Pequignot. JM; Favier. R. (2005). Age-related sensitivity to
lung oxidative stress during ozone exposure. Free Radic Res 39: 305-316.
http://dx.doi.org/10.1080/10715760400011098
Sherry. B; Blanck. HM; Galuska. DA; Pan. L; Dietz. WH; Balluz. L. (2010). Vital signs: State-specific obesity
prevalence among adults - United States, 2009. MMWR Recomm Rep 59: 951-955.
Shore. SA. (2007). Obesity and asthma: Lessons from animal models [Review]. J Appl Physiol 102: 516-528.
http://dx.doi.org/10.1152/iapplphysiol.00847.2006
Shore. SA; Lang. JE; Kasahara. PI; Lu. FL; Verbout. NG: Si. H; Williams. ES: Terry. RD; Lee. A; Johnston.
RA. (2009). Pulmonary responses to subacute ozone exposure in obese vs. lean mice. J Appl Physiol 107:
1445-1452. http://dx.doi.org/10.1152/iarjplphvsiol.00456.2009
Sienra-Monge. JJ: Ramirez-Aguilar. M; Moreno-Macias. H; Reyes-Ruiz. NI; Del Rio-Navarro. BE; Ruiz-
Navarro. MX; Hatch. G: Crissman. K; Slade. R: Devlin. RB; Romieu. I. (2004). Antioxidant
supplementation and nasal inflammatory responses among young asthmatics exposed to high levels of
ozone. Clin Exp Immunol 138: 317-322. http://dx.doi.Org/10.llll/i.1365-2249.2004.02606.x
8-46
-------
Silverman. RA; Ito. K. (2010). Age-related association of fine particles and ozone with severe acute asthma in
New York City. J Allergy Clin Immunol 125: 367-373. http://dx.doi.Org/10.1016/i.jaci.2009.10.061
SSDAN CensusScope (Social Science Data Analysis Network, CensusScope). (2010a). United States: Age
distribution [Database]. Ann Arbor, Michigan: Social Science Data Analysis Network. Retrieved from
http://www.censusscope.org/us/chart age.html
SSDAN CensusScope (Social Science Data Analysis Network, CensusScope). (2010b). United States:
Population by race [Database]. Ann Arbor, Michigan. Retrieved from
http://www.censusscope.org/us/chart race.html
SSDAN CensusScope (Social Science Data Analysis Network, CensusScope). (2010c). United States: Poverty
by age [Database]. Ann Arbor, Michigan. Retrieved from
http://www.censusscope.org/us/chart_povertv.html
Stafoggia. M; Forastiere. F; Faustini. A; Biggeri. A; Bisanti. L; Cadum. E; Cernigliaro. A; Mallone. S:
Pandolfi. P; Serinelli. M; Tessari. R; Vigotti. MA; Perucci. CA. (2010). Susceptibility factors to ozone-
related mortality: A population-based case-crossover analysis. Am J Respir Crit Care Med 182: 376-384.
httrj://dx.doi.org/10.1164/rccm.200908-1269OC
Steinvil. A; Kordova-Biezuner. L; Shapira. I; Berliner. S: Rogowski. O. (2008). Short-term exposure to air
pollution and inflammation-sensitive biomarkers. Environ Res 106: 51-61.
http://dx.doi.0rg/10.1016/i.envres.2007.08.006
Tankerslev. CG: Kleeberger. SR. (1994). Ozone-induced inflammation and altered ventilation in genetically
susceptible mice: A comparison of acute and subacute exposures. Toxicol Lett 72: 279-289.
Tankerslev. CG: Peng. RD; Bedga. D; Gabrielson. K; Champion. HC. (2010). Variation in echocardiographic
and cardiac hemodynamic effects of PM and ozone inhalation exposure in strains related to Nppa and Nprl
gene knock-out mice. Inhal Toxicol 22: 695-707. http://dx.doi.org/10.3109/08958378.2010.487549
Thaller. El; Petronella. SA; Hochman. D; Howard. S: Chhikara. RS: Brooks. EG. (2008). Moderate increases
in ambient PM2.5 and ozone are associated with lung function decreases in beach lifeguards. J Occup
Environ Med 50: 202-211. http://dx.doi.org/10.1097/JOM.Ob013e31816386b4
Torres. A; Utell. MJ; Morow. PE; Voter. KZ; Whitin. JC: Cox. C: Loonev. RJ: Speers. DM; Tsai. Y; Frampton.
MW. (1997). Airway inflammation in smokers and nonsmokers with varying responsiveness to ozone. Am
J Respir Crit Care Med 156: 728-736.
Tovalin. H; Valverde. M; Morandi. MT; Blanco. S: Whitehead. L; Rojas. E. (2006). DNA damage in outdoor
workers occupationally exposed to environmental air pollutants. Occup Environ Med 63: 230-236.
Trenga. CA; Koenig. JO; Williams. PV. (2001). Dietary antioxidants and ozone-induced bronchial
hyperresponsiveness in adults with asthma. Arch Environ Occup Health 56: 242-249.
U.S. Census Bureau. (2010). U.S. population projections [Database]. Retrieved from
http://www.census.gov/population/www/projections/projectionsagesex.html
U.S. EPA (U.S. Environmental Protection Agency). (1996a). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/P-93/004AF). Research Triangle Park,NC.
U.S. EPA (U.S. Environmental Protection Agency). (2006a). Aging and toxic response: Issues relevant to risk
assessment [EPA Report]. (EPA/600/P-03/004A). Washington, DC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=156648
U.S. EPA (U.S. Environmental Protection Agency). (2006b). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/R-05/004AF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
U.S. EPA (U.S. Environmental Protection Agency). (2009d). Integrated science assessment for paniculate
matter [EPA Report]. (EPA/600/R-08/139F). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=216546
8-47
-------
U.S. EPA (U.S. Environmental Protection Agency). (2010c). Integrated science assessment for carbon
monoxide [EPA Report]. (EPA/600/R-09/019F). Research Triangle Park,NC.
http ://cfpub. epa. gov/ncea/cfm/recordisplav. cfm?deid=218686
Vagaggini. B; Bartoli. MLE; Cianchetti. S: Costa. F; Bacci. E; Dente. FL; Di Franco. A; Malagrino. L;
Paggiaro. P. (2010). Increase in markers of airway inflammation after ozone exposure can be observed also
in stable treated asthmatics with minimal functional response to ozone. Respir Res 11:5.
http://dx.doi.org/10.1186/1465-9921-ll-5
Vagaggini. B; Taccola. M; Cianchetti. S: Carnevali. S: Bartoli. ML; Bacci. E; Dente. FL; Di Franco. A;
Giannini. D; Paggiaro. PL. (2002). Ozone exposure increases eosinophilic airway response induced by
previous allergen challenge. Am J Respir Crit Care Med 166: 1073-1077.
Valacchi. G: Pecorelli. A; Mencarelli. M; Maioli. E; Davis. PA. (2009). Beta-carotene prevents ozone-induced
proinflammatory markers in murine skin. Toxicol Ind Health 25: 241-247.
http://dx.doi.org/10.1177/0748233709103030
Valacchi. G: Vasu. VT; Yokohama. W: Corbacho. AM; Phung. A; Lim. Y; Aung. HH; Cross. CE; Davis. PA.
(2007). Lung vitamin E transport processes are affected by both age and environmental oxidants in mice.
Toxicol Appl Pharmacol 222: 227-234. http://dx.doi.Org/10.1016/i.taap.2007.04.010
Vancza. EM; Galdanes. K; Gunnison. A; Hatch. G: Gordon. T. (2009). Age, strain, and gender as factors for
increased sensitivity of the mouse lung to inhaled ozone. Toxicol Sci 107: 535-543.
http://dx.doi.org/10.1093/toxsci/kfn253
Villeneuve. PJ: Chen. L; Rowe. BH; Coates. F. (2007). Outdoor air pollution and emergency department visits
for asthma among children and adults: A case-crossover study in northern Alberta, Canada. Environ Health
Global Access Sci Source 6: 40. http://dx.doi.org/10.1186/1476-069X-6-40
Villeneuve. PJ: Chen. L; Stieb. D; Rowe. BH. (2006a). Associations between outdoor air pollution and
emergency department visits for stroke in Edmonton, Canada. Eur J Epidemiol 21: 689-700.
http://dx.doi.org/10.1007/sl0654-006-9050-9
Vovnow. JA; Fischer. BM; Zheng. S: Potts. EN; Grover. AR; Jaiswal. AK; Ohio. AJ; Foster. WM. (2009).
NAD(P)H quinone oxidoreductase 1 is essential for ozone-induced oxidative stress in mice and humans.
Am J Respir Cell Mol Biol 41: 107-113. http://dx.doi.org/10.1165/rcmb.2008-0381OC
Wagner. JG: Harkema. JR; Jiang. Q; Illek. B; Ames. BN; Peden. DB. (2009). Gamma-tocopherol attenuates
ozone-induced exacerbation of allergic rhinosinusitis in rats. Toxicol Pathol 37: 481-491.
http://dx.doi.org/10.1177/0192623309335630
Wagner. JG: Jiang. Q; Harkema. JR: Illek. B; Patel. DP; Ames. BN; Peden. DB. (2007). Ozone enhancement
of lower airway allergic inflammation is prevented by gamma-tocopherol. Free Radic Biol Med 43: 1176-
1188. http://dx.doi.0rg/10.1016/i.freeradbiomed.2007.07.013
Wattiez. R; Noel-Georis. I; Cruvt. C: Broeckaert. F; Bernard. A; Falmagne. P. (2003). Susceptibility to
oxidative stress: proteomic analysis of bronchoalveolar lavage from ozone-sensitive and ozone-resistant
strains of mice. Proteomics 3: 658-665. http://dx.doi.org/10.1002/pmic.200300417
Weinmann. GG: Weidenbach-Gerbase. M; Foster. WM; Zacur. H; Frank. R. (1995c). Evidence for ozone-
induced small-airway dysfunction: Lack of menstrual-cycle and gender effects. Am J Respir Crit Care Med
152: 988-996.
Williams. AS: Leung. SY; Nath. P; Khorasani. NM; Bhavsar. P; Issa. R: Mitchell. JA; Adcock. IM; Chung. KF.
(2007b). Role of TLR2, TLR4, and MyD88 in murine ozone-induced airway hyperresponsiveness and
neutrophilia. JAppl Physiol 103: 1189-1195. http://dx.doi.org/10.1152/japplphvsiol.00172.2007
Williams. AS: Nath. P; Leung. SY; Khorasani. N; Mckenzie. ANJ; Adcock. IM; Chung. KF. (2008b).
Modulation of ozone-induced airway hyperresponsiveness and inflammation by interleukin-13. Eur Respir
J 32: 571-578. http://dx.doi.org/10.1183/09031936.00121607
8-48
-------
Wong. CM; Ou. CO: Chan. KP; Chau. YK; Thach. TQ; Yang. L; Chung. RY; Thomas. GN; Peiris. JS: Wong.
TW: Hedlev. AJ; Lam. TH. (2008). The effects of air pollution on mortality in socially deprived urban areas
in Hong Kong, China. Environ Health Perspect 116: 1189-1194. http://dx.doi.org/10.1289/ehp.10850
Wong. CM: Yang. L: Thach. TO: Chau. PY: Chan. KP: Thomas. GN: Lam. TH: Wong. TW: Hedlev. AJ: Peiris.
JS. (2009). Modification by influenza on health effects of air pollution in Hong Kong. Environ Health
Perspect 117: 248-253. htrp://dx.doi.org/10.1289/ehp.H605
Yoon. HK; Cho. HY; Kleeberger. SR. (2007). Protective role of matrix metalloproteinase-9 in ozone-induced
airway inflammation. Environ Health Perspect 115: 1557-1563. http://dx.doi.org/10.1289/ehp.10289
Yu. M; Zheng. X; Witschi. H; Pinkerton. KE. (2002). The role of interleukin-6 in pulmonary inflammation and
injury induced by exposure to environmental air pollutants. Toxicol Sci 68: 488-497.
Zanobetti. A; Schwartz. J. (2008a). Is there adaptation in the ozone mortality relationship: A multi-city case-
crossover analysis. Environ Health 7: 22. http://dx.doi.org/10.1186/1476-069X-7-22
8-49
-------
9 ENVIRONMENTAL EFFECTS: OZONE EFFECTS ON
VEGETATION AND ECOSYSTEMS
9.1 Introduction
This chapter synthesizes and evaluates the relevant science to help form the scientific
foundation for the review of a vegetation- and ecologically-based secondary NAAQS
for O3. The secondary NAAQS are based on welfare effects. The Clean Air Act
(CAA) definition of welfare effects includes, but is not limited to, effects on soils,
water, wildlife, vegetation, visibility, weather, and climate, as well as effects on
materials, economic values, and personal comfort and well-being. The effects of O3
as a greenhouse gas and its direct effects on climate are discussed in Chapter 10 of
this document.
The intent of the ISA, according to the CAA, is to "accurately reflect the latest
scientific knowledge expected from the presence of [a] pollutant in ambient air" (42
U.S.C.7408 and 42 U.S.C.7409. This chapter of the ISA includes scientific research
from biogeochemistry, soil science, plant physiology, and ecology conducted at
multiple levels of biological organization (e.g., molecular, organ, organism,
population, community, ecosystem). Key information and judgments formerly found
in the AQCDs regarding O3 effects on vegetation and ecosystems are found in this
chapter. This chapter of the O3 ISA serves to update and revise Chapter 9 and AX9
of the 2006 O3 AQCD (U.S. EPA. 2006b).
Numerous studies of the effects of O3 on vegetation and ecosystems were reviewed
in the 2006 O3 AQCD. That document concluded that the effects of ambient O3 on
vegetation and ecosystems appear to be widespread across the U.S., and experimental
studies demonstrated plausible mechanisms for these effects. Ozone effect studies
published from 2005 to July 2011 are reviewed in this document in the context of the
previous O3 AQCDs. From 2005 to 2011, some areas have had very little new
research published and the reader is referred back to sections of the 2006 O3 AQCD
for a more comprehensive discussion of those subjects. This chapter is focused on
studies of vegetation and ecosystems that occur in the U.S. and that provide
information on endpoints or processes most relevant to the review of the secondary
standard. Many studies have been published about vegetation and ecosystems outside
of the U.S. and North America, largely in Europe and Asia. This document includes
discussion of studies of vegetation and ecosystems outside of North America only if
those studies contribute to the general understanding of O3 effects across species and
ecosystems. For example, studies outside North America are discussed that consider
physiological and biochemical processes that contribute to the understanding of
effects of O3 across species. Also, ecosystem studies outside of North America that
contribute to the understanding of O3 effects on general ecosystem processes are
discussed in the chapter.
9-1
-------
Sections of this chapter first discuss exposure methods, followed by effects on
vegetation and ecosystems at various levels of biological organization and ends with
policy-relevant discussions of exposure indices and exposure-response. Figure 9-1 is
a simplified illustrative diagram of the major endpoints O3 may affect. First, Section
9.2 presents a brief overview of various methodologies that have been, and continue
to be, central to quantifying O3 effects on vegetation (see AX9.1 of the 2006 O3
AQCD for more detailed discussion) (U.S. EPA. 2006b). Section 9.3 through Section
9.4 begin with a discussion of effects at the cellular and subcellular level followed by
consideration of the O3 effects on plant and ecosystem processes (Figure 9-1).
In Section 9.3 research is reviewed from the molecular to the biochemical and
physiological levels in impacted plants, offering insight into the mode of action of
O3. Section 9.4 provides a review of the effects of O3 exposure on major endpoints at
the whole plant scale including growth, reproduction, visible foliar injury and leaf
gas exchange in woody and herbaceous plants in the U.S., as well as a brief
discussion of O3 effects on agricultural crop yield and quality. Section 9.4 also
integrates the effects of O3 on individual plants in a discussion of available research
for assessing the effect of O3 on ecosystems, along with available studies that could
inform assessments of various ecosystem services (see Section 9.4.1.2).
The development of indices of O3 exposure and dose modeling is discussed in
Section 9.5. Finally, exposure-response relationships for a number of tree species,
native vegetation, and crop species and cultivars are reviewed, tabulated, and
compared in Section 9.6 to form the basis for an assessment of the potential risk to
vegetation from current ambient levels of O3.
9-2
-------
Effects of Ozone Exposure
Leaf metabolism & physiology
•Antioxidant metabolism up-regulated
•Decreased photosynthesis
•Decreased stomatal conductance
or sluggish stomatal response
Leaves & canopy
-Visible leaf injury
-Altered leaf senescence
•Altered leaf chemical composition
Plant growth (Fig 9-8)
•Decreased biomass accumulation
•Altered reproduction
•Altered carbon allocation
•Altered crop quality
Ecosystem services
•Decreased productivity
•Decreased C sequestration
•Altered water cycling (Fig 9-7)
•Altered community composition
(i.e., plant, insect & microbe)
Belowground processes (Fig 9-8)
•Altered litter production & decomposition
•Altered soil carbon & nutrient cycling
•Altered soil fauna & microbial communities
Figure 9-1 An illustrative diagram of the major endpoints that Os may affect in
plants and ecosystems.
9.2 Experimental Exposure Methodologies
9.2.1 Introduction
A variety of methods for studying plant response to O3 exposures have been
developed over the last several decades. The majority of methodologies currently
used have been discussed in detail in the 1996 O3 AQCD (U.S. EPA. 1996a) and
2006 O3 AQCD (U.S. EPA. 2006b). This section will serve as a short overview of
the methodologies and the reader is referred to the previous O3 AQCDs for more in-
depth discussion.
9-3
-------
9.2.2 "Indoor," Controlled Environment, and Greenhouse Chambers
The earliest experimental investigations of the effects of O3 on plants utilized simple
glass or plastic-covered chambers, often located within greenhouses, into which a
flow of O3-enriched air or oxygen could be passed to provide the exposure.
The types, shapes, styles, materials of construction, and locations of these chambers
have been numerous. Hogsett et al. (T987a) have summarized the construction and
performance of more elaborate and better instrumented chambers since the 1960s,
including those installed in greenhouses (with or without some control of
temperature and light intensity).
One greenhouse chamber approach that continues to yield useful information on the
relationships of O3 uptake to both physiological and growth effects employs
continuous stirred tank reactors (CSTRs) first described by Heck et al. (1978).
Although originally developed to permit mass-balance studies of O3 flux to plants,
their use has more recently widened to include short-term physiological and growth
studies of O3 * CO2 interactions (Loats and Rebbeck. 1999: Reinert et al.. 1997: Rao
et al.. 1995: Reinert and Ho. 1995: Heagle et al.. 1994a). and validation of visible
foliar injury on a variety of plant species (Kline et al.. 2009: Orendovici et al.. 2003).
In many cases, supplementary lighting and temperature control of the surrounding
structure have been used to control or modify the environmental conditions (Heagle
etal. 1994a).
Many investigations have utilized commercially available controlled environment
chambers and walk-in rooms adapted to permit the introduction of a flow of O3 into
the controlled air-volume. Such chambers continue to find use in genetic screening
and in physiological and biochemical studies aimed primarily at improving the
understanding of modes of action. For example, some of the studies of the O3
responses of common plantain (Plantago major) populations have been conducted in
controlled environment chambers (Whitfield et al.. 1996: Reiling and Davison.
1994).
More recently, some researchers have been interested in attempting to investigate
direct O3 effects on reproductive processes, separate from the effects on vegetative
processes (Black et al., 2010). For this purpose, controlled exposure systems have
been employed to expose the reproductive structures of annual plants to gaseous
pollutants independently of the vegetative component (Black et al., 2010: Stewart et
al.. 1996).
9.2.3 Field Chambers
In general, field chamber studies are dominated by the use of various versions of the
open top chamber (OTC) design, first described by Heagle et al. (1973) and Mandl et
al. (1973). The OTC method continues to be a widely used technique in the U.S. and
Europe for exposing plants to varying levels of O3. Most of the new information
confirms earlier conclusions and provides additional support for OTC use in
9-4
-------
assessing plant species and in developing exposure-response relationships. Chambers
are generally ~3 meters in diameter with 2.5 meter-high walls. Hogsett et al. (T987b)
described in detail many of the various modifications to the original OTC designs
that appeared subsequently, e.g., the use of larger chambers for exposing small trees
(Kats et al., 1985) or grapevines (Mandl et al., 1989), the addition of a conical baffle
at the top to improve ventilation (Kats et al.. 1976). a frustum at the top to reduce
ambient air incursions, and a plastic rain-cap to exclude precipitation (Hogsett et al..
1985). All versions of OTCs included the discharge of air via ports in annular
ducting or interiorly perforated double-layered walls at the base of the chambers to
provide turbulent mixing and the upward mass flow of air.
Chambered systems, including OTCs, have several advantages. For instance, they
can provide a range of treatment levels including charcoal-filtered (CF), clean-air
control, and several above ambient concentrations for O3 experiments. Depending on
experimental intent, a replicated, clean-air control treatment is an essential
component in many experimental designs. The OTC can provide a consistent,
definable exposure because of the constant wind speed and delivery systems.
Statistically robust concentration-response (C-R) functions can be developed using
such systems for evaluating the implications of various alternative air quality
scenarios on vegetation response. Nonetheless, there are several characteristics of the
OTC design and operation that can lead to exposures that might differ from those
experienced by plants in the field. First, the OTC plants are subjected to constant air
flow turbulence, which, by lowering the boundary layer resistance to diffusion, may
result in increased uptake. This may lead to an overestimation of effects relative to
areas with less turbulence (Krupa et al.. 1995; Legge et al.. 1995). However, other
research has found that OTC's may slightly change vapor pressure deficit (VPD) in a
way that may decrease the uptake of O3 into leaves (Piikki et al.. 2008a). As with all
methods that expose vegetation to modified O3 concentrations in chambers, OTCs
create internal environments that differ from ambient air. This so-called "chamber
effect" refers to the modification of microclimatic variables, including reduced and
uneven light intensity, uneven rainfall, constant wind speed, reduced dew formation,
and increased air temperatures (Fuhrer. 1994: Manning and Krupa. 1992). However,
in at least one case where canopy resistance was quantified in OTCs and in the field,
it was determined that gaseous pollutant exposure to crops in OTCs was similar to
that which would have occurred at the same concentration in the field (TJnsworth et
al.. 1984a. b). Because of the standardized methodology and protocols used in
National Crop Loss Assessment Network (NCLAN) and other programs, the
database can be assumed to be internally consistent.
While it is clear that OTCs can alter some aspects of the microenvironment and plant
growth, it is important to establish whether or not these differences affect the relative
response of a plant to O3. As noted in the 1996 O3 AQCD (U.S. EPA. 1996a).
evidence from a number of comparative studies of OTCs and other exposure systems
suggested that responses were essentially the same regardless of exposure system
used and chamber effects did not significantly affect response. In studies that
included exposure to ambient concentrations of O3 in both OTCs, and open-air,
chamberless control plots, responses in the OTCs were the same as in open-air plots.
9-5
-------
Examples include studies of tolerant and sensitive white clover clones (Trifolium
repens) to ambient O3 in greenhouse, open top, and ambient plots (Heagle et al,
1996), Black Cherry (Primus serotind) (Neufeld et al., 1995), and three species of
conifers (Neufeld et al., 2000). Experimental comparisons between exposure
methodologies are reviewed in Section 9.2.6.
Another type of field chamber called a "terracosm" has been developed and used in
recent studies (Lee et al., 2009a). Concern over the need to establish realistic plant-
litter-soil relationships as a prerequisite to studies of the effects of O3 and CO2
enrichment on ponderosa pine (Pinus ponderosd) seedlings led Tingey et al. (1996)
to develop closed, partially environmentally controlled, sun-lit chambers
("terracosms") incorporating lysimeters (1 meter deep) containing forest soil in
which the appropriate horizon structure was retained.
Other researchers have recently published studies using another type of out-door
chamber called recirculating Outdoor Plant Environment Chambers (OPECs)
(Flowers et al.. 2007). These closed chambers are approximately 2.44 meters xi.52
meters with a growth volume of approximately 3.7 m3 in each chamber. These
chambers admit 90% of full sunlight and control temperature, humidity and vapor
pressure (Fiscus et al.. 1999).
9.2.4 Plume and FACE-Type Systems
Plume systems are chamberless exposure facilities in which the atmosphere
surrounding plants in the field is modified by the injection of pollutant gas into the
air above or around them from multiple orifices spaced to permit diffusion and
turbulence, so as to establish relatively homogeneous conditions as the individual
plumes disperse and mix with the ambient air. They can only be used to increase the
O3 levels in the ambient air.
The most common plume system used in the U.S. is a modification of the free-air
carbon dioxide/ozone enrichment (FACE) system (Hendrey et al., 1999; Hendrey and
Kimball, 1994). Although originally designed to provide chamberless field facilities
for studying the CO2 effects of climate change, FACE systems have been adapted to
include the dispensing of O3 (Karnosky et al., 1999). This method has been
employed in Illinois (SoyFACE) to study soybeans (Morgan et al., 2004; Rogers et
al., 2004) and in Wisconsin (Aspen FACE) to study trembling aspen (Populus
tremuloides), birch (Betula papyri/era) and maple (Acer saccharum) (Karnosky et
al., 1999). Volk et al. (2003) described a similar system for exposing grasslands that
uses 7-m diameter plots. Another similar FACE system has been used in Finland
(SavirantaetaL2010; Oksanen. 2003).
The FACE systems in the U.S. discharge the pollutant gas (O3 and/or CO2) through
orifices spaced along an annular ring (or torus) or at different heights on a ring of
vertical pipes. Computer-controlled feedback from the monitoring of gas
concentration regulates the feed rate of enriched air to the dispersion pipes. Feedback
9-6
-------
of wind speed and directional information ensures that the discharges only occur
upwind of the treatment plots, and that discharge is restricted or closed down during
periods of low wind speed or calm conditions. The diameter of the arrays and their
height (25-30 meters) in some FACE systems requires large throughputs of enriched
air per plot, particularly in forest tree systems. The cost of the throughputs tends to
limit the number of enrichment treatments, although Hendrey et al. (1999) argued
that the cost on an enriched volume basis is comparable to that of chamber systems.
A different FACE-type facility has been developed for the Kranzberg Ozone
Fumigation Experiment (KROFEX) in Germany beginning in 2000 (Nunn et al.,
2002; Werner and Fabian, 2002). The experiment aims to study the effects of O3 on
mature stands of beech (Fagus sylvatica) and spruce (Picea abies) trees in a system
that functions independently of wind direction. The enrichment of a large volume of
the ambient air immediately above the canopy takes place via orifices in vertical
tubes suspended from a horizontal grid supported above the canopy.
Although plume systems make virtually none of the modifications to the physical
environment that are inevitable with chambers, their successful use depends on
selecting the appropriate numbers, sizes, and orientations of the discharge orifices to
avoid "hot-spots" resulting from the direct impingement of jets of pollutant-enriched
air on plant foliage (Werner and Fabian. 2002). Because mixing is unassisted and
completely dependent on wind turbulence and diffusion, local gradients are
inevitable especially in large-scale systems. FACE systems have provisions for
shutting down under low wind speed or calm conditions and for an experimental area
that is usually defined within a generous border in order to strive for homogeneity of
the exposure concentrations within the treatment area. They are also dependent upon
continuous computer-controlled feedback of the O3 concentrations in the mixed
treated air and of the meteorological conditions. Plume and FACE systems also are
unable to reduce O3 levels below ambient in areas where O3 concentrations are
phytotoxic.
9.2.5 Ambient Gradients
Ambient O3 gradients that occur in the U.S. hold potential for the examination of
plant responses over multiple levels of exposure. However, few such gradients can be
found that meet the rigorous statistical requirements for comparable site
characteristics such as soil type, temperature, rainfall, radiation, and aspect (Manning
andKrupa. 1992): although with small plants, soil variability can be avoided by the
use of plants in large pots. The use of soil monoliths transported to various locations
along natural O3 gradients is another possible approach to overcome differences in
soils; however, this approach is also limited to small plants.
Studies in the 1970s used the natural gradients occurring in southern California to
assess yield losses of alfalfa and tomato (Oshima et al., 1977; Oshima et al., 1976).
A transect study of the impact of O3 on the growth of white clover and barley in the
U.K. was confounded by differences in the concurrent gradients of SO2 and NO2
9-7
-------
pollution (Ashmore et al., 1988). Studies of forest tree species in national parks in the
eastern U.S. (Winner et al., 1989) revealed increasing gradients of O3 and visible
foliar injury with increased elevation.
Several studies have used the San Bernardino Mountains Gradient Study in southern
California to study the effects of O3 and N deposition on forests dominated by
ponderosa and Jeffrey pine (Jones and Paine, 2006; Arbaugh et al., 2003; Grulke,
1999; U.S. EPA, 1977). However, it is difficult to separate the effects of N and O3 in
some instances in these studies (Arbaugh et al., 2003). An O3 gradient in Wisconsin
has been used to study foliar injury in a series of trembling aspen clones (Populus
tremuloides) differing in O3 sensitivity (Mankovska et al., 2005; Karnosky et al.,
1999). Also in the Midwest, an east-west O3 gradient around southern Lake
Michigan was used to look at growth and visible foliar injury in (P. serotina) and
common milkweed (Asclepias syriacd) (Bennett et al., 2006).
More recently, studies have been published that have used natural gradients to study
a variety of endpoints and species. For example, Gregg et al. (2003) studied
cottonwood (Populus deltoides) saplings grown in an urban to rural gradient of O3 by
using seven locations in the New York City area. The secondary nature of the
reactions of O3 formation and NOX titration reactions within the city center resulted
in significantly higher cumulative O3 exposures in more rural sites. Potential
modifying factors such as soil composition, moisture, or temperature were either
controlled or accounted for in analysis. As shown in Section 9.6.3.3. the response of
this species to O3 exposure was much stronger than most species. The natural
gradient exposures were reproduced in parallel using OTCs, and yielded similar
results. Also, the U.S. Forest Service - Forest Inventory and Analysis (FIA) program
uses large-scale O3 exposure patterns across the continental U.S. to study
occurrences of foliar injury due to O3 exposure (Smith et al., 2003) (Section 9.4.2).
Finally, McLaughlin et al. (2007a; 2007b) used spatial and temporal O3 gradients to
study forest growth and water use in the southern Appalachians. These studies found
varying O3 exposures between years and between sites.
9.2.6 Comparative Studies
All experimental approaches used to expose plants to O3 have strengths and
weaknesses. One potential weakness of laboratory, greenhouse, or field chamber
studies is the potential effect of the chamber on micrometeorology. In contrast,
plume, FACE and gradient systems are limited by the very small number of possible
exposure levels (almost always no more than two), small replication and the inability
to reduce O3 levels below ambient. In general, experiments that aim at characterizing
the effect of a single variable, e.g., exposure to O3, must not only manipulate the
levels of that variable, but also control potentially interacting variables and
confounders, or else account for them. However, while increasing control of
environmental variables makes it easier to discern the effect of the variable of
interest, it must be balanced with the ability to extend conclusions to natural,
9-8
-------
non-experimental settings. More naturalistic exposure systems, on the other hand, let
interacting factors vary freely, resulting in greater unexplainable variability.
The various exposure methodologies used with O3 vary in the balance each strikes
between control of environmental inputs, closeness to the natural environment,
noisiness of the response data, and ability to make general inferences.
Studies have examined the comparability of results obtained though the various
exposure methodologies. As noted in the 1996 O3 AQCD, evidence from the
comparative studies of OTCs and from closed chamber and O3-exclusion exposure
systems on the growth of alfalfa (Medicago saliva) by Olszyk et al. (1986) suggested
that, since significant differences were found for fewer than 10% of the growth
parameters measured, the responses were, in general, essentially the same regardless
of exposure system used, and chamber effects did not significantly affect response.
Heagle et al. (1988) concluded: "Although chamber effects on yield are common,
there are no results showing that this will result in a changed yield response to O3."
A study of the effects of an enclosure examined the responses of tolerant and
sensitive white clover clones (Trifolium repens) to ambient O3 in a greenhouse,
open-top chamber, and ambient (no chamber) plots (Heagle et al., 1996). For
individual harvests, greenhouse O3 exposure reduced the forage weight of the
sensitive clone 7 to 23% more than in OTCs. However, the response in OTCs was
the same as in ambient plots. Several studies have shown very similar response of
yield to O3 for plants grown in pots or in the ground, suggesting that even such a
significant change in environment does not alter the proportional response to O3,
providing that the plants are well watered (Heagle et al., 1983; Heagle, 1979).
A few recent studies have compared results of O3 experiments between OTCs, FACE
experiments, and gradient studies. For example, a series of studies undertaken at
Aspen FACE (Isebrands et al., 2001; Isebrands et al., 2000) showed that O3 symptom
expression was generally similar in OTCs, FACE, and ambient O3 gradient sites, and
supported the previously observed variation among trembling aspen clones using
OTCs (Mankovska et al., 2005; Karnosky et al., 1999). In the SoyFACE experiment
in Illinois, soybean (Pioneer 93B15 cultivar) yield loss data from a two-year study
was published (Morgan et al., 2006). This cultivar is a recent selection and, like most
modern cultivars, has been selected under an already high current O3 exposure.
It was found to have average sensitivity to O3 compared to 22 other cultivars tested
at SoyFACE. In this experiment, ambient hourly O3 concentrations were increased
by approximately 20% and measured yields were decreased by 15% in 2002 as a
result of the increased O3 exposure (Morgan et al., 2006). To compare these results
to chamber studies, Morgan et al. (2006) calculated the expected yield loss from a
linear relationship constructed from chamber data using seven-hour seasonal
averages (Ashmore, 2002). They calculated an 8% expected yield loss from the 2002
O3 exposure using that linear relationship. As reported in Section 9.2.5, Gregg et al.
(2006, 2003) found similar O3 effects on cottonwood sapling biomass growth along
an ambient O3 gradient in the New York City area and a parallel OTC study.
Finally, EPA conducted comparisons of exposure-response model predictions based
on OTC studies, and more recent FACE observations. These comparisons include
9-9
-------
yield of annual crops, and biomass growth of trees. They are presented in Section
9.6.3 of this document.
9.3 Mechanisms Governing Vegetation Response to Ozone
9.3.1 Introduction
This section focuses on the effects of O3 stress on plants and their responses to that
stress on the molecular, biochemical and physiological levels. First, the pathway of
O3 uptake into the leaf and the initial chemical reactions occurring in the substomatal
cavity and apoplast will be described (Section 9.3.2): additionally, direct effects of
O3 on the stomatal apparatus will be discussed. Once O3 has entered the substomatal
cavity and apoplast, the cell initiates rapid changes in signaling pathways and gene
expression that have been measured in O3-treated plants. The next section focuses on
changes in gene and protein expression measured in plants exposed to O3, with
particular emphasis on results from transcriptome (all RNA molecules produced in a
cell) and proteome (all proteins produced in a cell) analyses (Section 9.3.3.2).
Subsequently, the role of phytohormones such as salicylic acid (SA), ethylene (ET),
jasmonic acid (JA), and abscisic acid (ABA) and their interactions in both signal
transduction processes and in determining plant response to O3 is discussed in
Section 9.3.3.3. After O3 uptake, some plants can respond to the oxidative stress with
detoxification to minimize damage. These mechanisms of detoxification, with
particular emphasis on antioxidant enzymes and metabolites, are reviewed in Section
9.3.4. The next section focuses on changes in primary and secondary metabolism in
plants exposed to O3, looking at photosynthesis, respiration and several secondary
metabolites, some of which may also act as antioxidants and protect the plant from
oxi dative stress (Section 9.3.5). For many of these topics, information from the 2006
O3 AQCD (U.S. EPA, 2006b) has been summarized, as this information is still valid
and supported by more recent findings. For other topics, such as genomics and
proteomics, which have arisen due to the availability of new technologies, the
information is based solely on new publications with no reference to the 2006 O3
AQCD.
As Section 9.3 focuses on mechanisms underlying effects of O3 on plants and their
response to it, the conditions that are used to study these mechanisms do not always
reflect conditions that a plant may be exposed to in an agricultural setting or natural
ecosystem. The goal of many of these studies is to generate an O3 effect in a
relatively short period of time and not always to simulate ambient O3 exposures.
Therefore, plants are often exposed to unrealistically high O3 concentrations for
several hours or days (acute exposure), and only in some cases to ambient or slightly
elevated O3 concentrations for longer time periods (chronic exposure). Additionally,
the plant species utilized in these studies are often not agriculturally important or
commonly found as part of natural ecosystems. Model organisms such as
Arabidopsis thaliana are used frequently as they are easy to work with, and mutants
9-10
-------
or transgenic plants are easy to develop or have already been developed.
Furthermore, the Arabidopsis genome has been sequenced, and much is known about
the molecular basis of many biochemical and cellular processes.
Many of the studies described in this section focus on changes in the expression of
genes in O3-treated plants. Some very recent studies utilizing proteomics techniques
have evaluated changes in protein expression for large numbers of proteins in O3
treated plants, and the findings from these studies support the previous results
regarding changes in gene expression studies as a result of O3 exposure. The next
step in the process is to determine the implications of the measured changes
occurring at the cellular level to whole plants and ecosystems, which is an important
topic of study which has not been widely addressed.
The most noteworthy new body of research since the 2006 O3 AQCD is on the
understanding of molecular mechanisms underlying how plants are affected by O3;
many of the recent studies reviewed here focus on changes in gene expression in
plants exposed to elevated O3. The findings summarized in the 2006 O3 AQCD
included decreases in transcript levels of photosynthesis associated genes, and
increases in transcript levels of genes encoding for pathogenesis-related proteins,
enzymes needed for ethylene synthesis, antioxidant enzymes and defense genes such
as phenylalanine ammonia lyase in plants exposed to O3. These findings have been
supported by the new studies, and the advent of new technologies has allowed for a
more comprehensive understanding of the mechanisms governing how pi ants are
affected by O3.
In summary, these new studies have increased knowledge of the molecular,
biochemical and cellular mechanisms occurring in plants in response to O3 by often
using artificial exposure conditions and model organisms. This information adds to
the understanding of the basic biology of how plants are affected by oxidative stress
in the absence of any other potential stressors. The results of these studies provide
important insights, even though they may not always directly translate into effects
observed in other plants under more realistic exposure conditions.
9.3.2 Ozone Uptake into the Leaf
Appendix AX9.2.3 of the 2006 O3 AQCD clearly described the process by which O3
enters plant leaves through open stomata (U.S. EPA, 2006b). This information
continues to be valid and is only summarized here.
Ozone moves from the atmosphere above the canopy boundary layer into the canopy
primarily by turbulent air flow. Canopy conductance is controlled by the complexity
of the canopy architecture. Within the canopy, O3 is adsorbed onto surfaces as well
as being absorbed into the leaves. Absorption into leaves is controlled by leaf
boundary layer and stomatal conductance, which together determine leaf
conductance (Figure 9-2. Panel A). Other factors that may also limit uptake include
the size of the stomatal aperture and the reactions of O3 with biogenically-emitted
9-11
-------
hydrocarbons (U.S. EPA, 2006b; Kurpius and Goldstein, 2003). Stomata provide the
principal pathway for O3 to enter and affect plants (Massman and Grantz. 1995;
Fuentes et al.. 1992; Reich, 1987; Leuning et al.. 1979). Ozone moves into the leaf
interior by diffusing through open stomata, and environmental conditions which
promote high rates of gas exchange will favor the uptake of the pollutant by the leaf
(Figure 9-2. Panel B). Once inside the substomatal cavity, O3 is thought to rapidly
react with the aqueous apoplast to form breakdown products known as reactive
oxygen species (ROS), such as hydrogen peroxide (H2O2), superoxide (O2~),
hydroxyl radicals (HO ) and peroxy radicals (HO2*) (Figure 9-3). Hydrogen peroxide
is not only a toxic breakdown product of O3, but has been shown to function as a
signaling molecule, which is activated in response to both biotic and abiotic stressors.
The role of H2O2 in signaling was described in detail in the 2006 O3 AQCD.
Additional organic molecules present in the apoplast or cell wall, such as those
containing double bonds or sulfhydryls that are sensitive to oxidation, could also be
converted to oxygenated molecules after interacting with O3 (Figure 9-4). These
reactions are not only pH dependent, but are also influenced by the presence of other
molecules in the apoplast (U.S. EPA. 2006b). The 2006 O3 AQCD provided a
comprehensive summary of these possible interactions of O3 with other biomolecules
(U.S. EPA. 2006b). It is in the apoplast that initial detoxification reactions by
antioxidant metabolites and enzymes take place, and these initial reactions are critical
to reduce concentrations of the oxidative breakdown products of O3; these reactions
are described in more detail in Section 9.3.4 of this document.
9.3.2.1 Changes in Stomatal Function
Ozone-induced changes in stomatal conductance have been reviewed in detail in
previous O3 AQCDs. The findings summarized in these documents demonstrate that
stomatal conductance is often reduced in plants exposed to O3, resulting either from
a direct impact of O3 on the stomatal complex which causes closure, or as a response
to increasing CO2 concentrations in the substomatal cavity as carbon fixation is
reduced. Although the nature of these effects depends upon many different factors,
including the plant species, concentration and duration of the O3 exposure, and
prevailing meteorological conditions, stomatal conductance is often negatively
affected by plant exposure to O3 (Wittig et al.. 2007). Decreases in conductance have
been shown to result from direct as well as indirect effects on stomata (Wittig et al..
2007). However, some recent studies have reported increased conductance in
response to O3 exposure, suggesting partial stomatal dysfunction (Paoletti and
Grulke. 2010).
Results from the use of Arabidopsis mutants and new technologies, which allow for
analysis of guard cell function in whole plants rather than in isolated guard cells or
epidermal peels, suggest that O3 may also have a direct impact on stomatal guard
cells, leading to alterations in stomatal conductance. The use of a new simultaneous
0.3. exposure/gas exchange device has demonstrated that exposure of Arabidopsis
ecotypes Col-0 and Ler to 150 ppb O3 resulted in a 60-70% decline in stomatal
9-12
-------
conductance within 9-12 minutes of beginning the exposure. Twenty to thirty
minutes later, stomatal conductance had returned to its initial value, even with
continuing exposure to O3, indicating a rapid direct effect of O3 on stomatal function
(Kollist et al., 2007). This transient decrease in stomatal conductance was not
observed in the abscisic acid insensitive (ABI2) Arabidopsis mutant. As the ABI2
protein is thought to regulate the signal transduction process involved in stomatal
response downstream of ROS production, the authors suggest that the transient
decrease in stomatal conductance in the Col-0 and Ler ecotypes results from the
biological action of ROS in transducing signals, rather than direct physical damage to
guard cells by ROS (Kollist et al.. 2007). This rapid transient decrease in stomatal
conductance was also not observed when exposing the Arabidopsis mutant slacl
(slow anion channel-associated 1) to 200 ppb O3 (Vahisalu et al.. 2008). The SLAC1
protein was shown to be essential for guard cell slow anion channel functioning and
for stomatal closure in response to O3. Based on additional studies using a variety of
Arabidopsis mutants impaired in various aspects of stomatal function, Vahisalu et al.
(2008) suggest that the presence of ROS in the guard cell apoplast (formed either by
O3 breakdown or through ROS production from NADPH oxidase activity) leads to
the activation of a signaling pathway in the guard cells, which includes SLAC1, and
results in stomatal closure.
A review by McAinsh et al. (2002) discusses the role of calcium as a part of the
signal transduction pathway involved in regulating stomatal responses to pollutant
stress. A number of studies in this review provide some evidence that exposure to O3
increases the cytosolic free calcium concentration ([Ca2+]cyt) in guard cells, which
may result in an inhibition of the plasma membrane inward-rectifying K+ channels in
guard cells, which allow for the K+ uptake needed for stomatal opening (McAinsh et
al.. 2002: Torsethaugen et al.. 1999). This would compromise the ability of the
stomata to respond to various stimuli, including light, CO2 concentration and
drought. Pei et al. (2000) reported that the presence of H2O2 activated Ca2+ -
permeable channels, which mediate increases in [Ca2+]cyt in guard cell plasma
membranes of Arabidopsis. They also determined that abscisic acid (ABA) induced
H2O2 production in guard cells, leading to ABA-induced stomatal closure via
activation of the membrane Ca2+ channels. Therefore, it is possible that H2O2, a
byproduct of O3 breakdown in the apoplast, could disrupt the Ca2+-ABA signaling
pathway that is involved in regulating stomatal responses (McAinsh et al.. 2002).
The studies described here provide some evidence to suggest that O3 and its
breakdown products can directly affect stomatal functioning by impacting the signal
transduction pathways which regulate guard cells. Stomatal sluggishness has been
described as a delay in stomatal response to changing environmental conditions in
sensitive species exposed to higher concentrations and/or longer-term O3 exposures
(Paoletti and Grulke. 2010. 2005: McAinsh et al.. 2002). It is possible that the
signaling pathways described above could be involved in mediating this stomatal
sluggishness in some plant species under certain O3 exposure conditions (Paoletti
and Grulke. 2005: McAinsh et al.. 2002).
9-13
-------
A.
B.
Light
Cuticle
Epidermis
Pallisade
Mesophyll
Spongy
Mesophyll
Epidermis
Cuticle1
II
Vascular
System
C0=[C02]--1
Note: While details among species vary, the general overview remains the same. Light that drives photosynthesis generally falls
upon the upper (adaxial) leaf surface. Carbon dioxide (CO) and O3 enter through the stomata, while water vapor exits through the
stomata (transpiration). Stomata are usually on the lower (abaxial) leaf surface, but may occur on the upper leaf surface in some
species.
Figure 9-2 Ozone uptake from the atmosphere (A), and The anatomy of a dicot
leaf (B).
9-14
-------
3
Radical
b.
HO,
Superoxide
!
t
H0
|H,0
Hydrags/i
PeNKKte
HO H20
2U2
Peroxy/
Note: (a) Ozone reacts at the double bonds to form carbonyl groups, (b) Under certain circumstances, peroxides are generated.
Figure 9-3 Possible reactions of O3 within water.
a.
H2C = CH2
Crigee
Mechanism / \
- »> 9 P
H2C - CH2
H
HOO
O
HC-OH
2.
3. NO,
H2C - CH2
ON02
H2C - CH2
b.
O=< >CH(OH)CH 02H
HO OH
HO OH
O O
\ /
O
CH(OH)CH O2H
•H202
HO OH
/ \
0=C CH(OH)CH02H
=cx /-
CHO CHO
Further Oxidation
Note: (a) The typical Crigee mechanism is shown in which several reaction paths from the initial product are shown, (b) Typical
reaction of ascorbic acid with O3.
Source: Adapted from Mudd et al. (1996).
Figure 9-4 The Crigee mechanism of O$ attack of a double bond.
9-15
-------
9.3.3 Cellular to Systemic Responses
9.3.3.1 Ozone Signal Transduction
New technologies allowing for large-scale analysis of oxidative stress-induced
changes in gene expression have facilitated the study of signal transduction processes
associated with plant response to O3 exposure. Many of these studies have been
conducted using Arabidopsis or tobacco plants, for which a variety of mutants are
available and/or which can be easily genetically modified to generate either loss-of-
function or over-expressing genotypes. Several comprehensive review articles
provide an overview of what is known of O3-induced signal transduction processes
and how they may help to explain differential sensitivity of plants to the pollutant
(Ludwikow and Sadowski, 2008; Baier et al., 2005; Kangasjarvi et al., 2005).
Additionally, analysis of several studies of transcriptome changes has also allowed
for the compilation of these data to determine an initial time-course for O3-induced
activation of various signaling compounds (Kangasjarvi et al., 2005).
Some of the earliest events that occur in plants exposed to O3 have been described in
the guard cells of stomata. Reactive oxygen species were observed in the chloroplasts
of guard cells in the O3 tolerant Col-0 Arabidopsis thaliana ecotype plants within
5 minutes of plant exposure to 350 ppb O3 (Joo et al., 2005). Reactive oxygen
species from the breakdown of O3 in the apoplast are believed to activate GTPases
(G-proteins), which, in turn, activate several intracellular sources of ROS, including
ROS derived from the chloroplasts. G-proteins are also believed to play a role in
activating membrane-bound NADPH oxidases to produce ROS and, as a result,
propagate the oxidative burst to neighboring cells (Joo et al., 2005). Therefore, G-
proteins are recognized as important molecules involved in plant responses to O3 and
may play a role in the initiation of signal transduction mechanisms resulting from the
presence of ROS in the apoplast (Kangasjarvi et al., 2005; Booker et al., 2004a).
A change in the redox state of the plant may contribute to initiating the process of
signaling of oxidative stress in plants. Disulfide-thiol conversions in proteins and the
redox state of the glutathione pool may be important components of redox and signal
transduction (Foyer and Noctor. 2005a. b).
Calcium (Ca2+) has also been implicated in the transduction of signals to the nucleus
in response to oxidative stress. The influx of Ca2+ from the apoplast into the cell
occurs early during plant exposure to O3, and it is thought to play a role in regulating
the activity of protein kinases, which are discussed below (Baier et al.. 2005; Hamel
et al.. 2005). Calcium channel blockers inhibited O3-induced activation of protein
kinases in tobacco suspension cells exposed to 500 ppb O3 for 10 minutes, indicating
that the opening of Ca2+ channels is an important upstream signaling event or that the
(as yet unknown) upstream process has a requirement for Ca2+ (Samuel et al.. 2000).
Further transmission of information regarding the presence of ROS to the nucleus
involves mitogen-activated protein kinases (MAPKs), which phosphorylate proteins
9-16
-------
and activate various cellular responses (Hamel et al.. 2005). Mitogen-activated
protein kinases are induced in several different plant species in response to O3
exposure, including tobacco (Samuel et al., 2005), Arabidopsis (Ludwikowet al.,
2004), the shrub Phillyrea latifolia (Paolacci et al., 2007) and poplar (Hamel et al.,
2005). Disruption of these signal transduction pathways by over-expressing or
suppressing MAPK activity in different Arabidopsis and tobacco lines resulted in
increased plant sensitivity to O3 (Miles et al.. 2005: Samuel and Ellis. 2002).
Additionally, greater O3 tolerance of several Arabidopsis ecotypes was correlated
with greater upregulation of MAPK signaling pathways upon O3 exposure than in
more sensitive Arabidopsis ecotypes (Li et al.. 2006b: Mahalingam et al.. 2006:
Overmyer et al.. 2005). indicating that determination of plant sensitivity and plant
response to O3 may, in part, be determined not only by whether these pathways are
turned on, but also by the magnitude of the signals moving through these
communication channels.
In conclusion, experimental evidence suggests that there are likely several different
mechanisms by which signaling as part of plant response to O3 or its breakdown
products is initiated. These mechanisms may vary by species or developmental stage
of the plant, or may co-exist and be activated by different exposure conditions.
Calcium and protein kinases are likely involved in relaying information to the
nucleus and other cellular compartments as a first step in determining whether and
how the plant will respond to the stress.
9.3.3.2 Gene and Protein Expression Changes in Response to
Ozone
The advent of DNA microarray technology has allowed for the study of gene
expression in cells on a large scale. Rather than assessing changes in gene expression
of individual genes, DNA microarrays facilitate the evaluation of entire
transcriptomes, providing a comprehensive picture of simultaneous alterations in
gene expression. In addition, these studies have provided more insight into the
complex interactions between molecules, how those interactions lead to the
communication of information in the cell (or between neighboring cells), and which
role these interactions play in determining tolerance or sensitivity and how a plant
may respond to stresses such as O3 (Ludwikow and Sadowski. 2008). Transcriptome
analysis of O3-treated plants has been performed in several species, including
Arabidopsis thaliana (Li et al.. 2006b: Tosti et al.. 2006: Heidenreich et al.. 2005:
Mahalingam et al.. 2005: Tamaoki et al.. 2003). pepper (Capsicum annuum) (Lee and
Yun. 2006). clover (Medicago truncatuld) (Puckette et al.. 2008). Phillyrea latifolia
(Paolacci et al.. 2007). poplar (Street et al.. 2011). and European beech (Fagus
sylvaticd) (Olbrich et al.. 2010: Olbrich et al.. 2009: Olbrich et al.. 2005). In some
cases, researchers compared transcriptomes of two or more cultivars, ecotypes or
mutants that differed in their sensitivity to O3 (Puckette et al.. 2008: Rizzo et al..
2007: Lee and Yun. 2006: Li et al.. 2006b: Tamaoki et al.. 2003). Species, O3
exposure conditions (concentration, duration of exposure) and sampling times varied
9-17
-------
considerably in these studies. However, functional classification of the genes that
were either upregulated or downregulated by plant exposure to O3 exhibited common
trends. Genes involved in plant defense, signaling and those associated with the
synthesis of plant hormones and secondary metabolism were generally upregulated,
while those related to photosynthesis and general metabolism were typically
downregulated in O3-treated plants (Puckette et al.. 2008: Lee and Yun. 2006: Li et
al.. 2006b: Tosti et al.. 2006: Olbrich et al.. 2005: Tamaoki et al. 2003).
Analysis of the transcriptome has been used to evaluate differences in gene
expression between sensitive and tolerant plants in response to O3 exposure.
In pepper, 67% of the 180 genes studied that were affected by O3 were differentially
regulated in the sensitive and tolerant cultivars. At both 0 hours and 48 hours after a
3-day exposure at 150 ppb, O3 responsive genes were either upregulated or
downregulated more markedly in the sensitive than in the tolerant cultivar (Lee and
Yun, 2006). Transcriptome analysis also revealed differences in timing and
magnitude of changes in gene expression between sensitive and tolerant clovers.
Acute exposure (300 ppb O3 for 6 hours) led to the production of an oxidative burst
in both clovers (Puckette et al., 2008). However, the sensitive-Jemalong cultivar
exhibited a sustained ROS burst and a concomitant downregulation of defense
response genes at 12 hours after the onset of exposure, while the tolerant JE 154
accession showed much more rapid and large-scale transcriptome changes than the
Jemalong cultivar (Puckette et al., 2008).
Arabidopsis ecotypes WS and Col-0 were exposed to 1.2 x ambient O3
concentrations for 8-12 days at the SoyFACE site (Li et al.. 2006b). The sensitive
WS ecotype showed a far greater number of changes in gene expression in response
to this low-level O3 exposure than the tolerant Col-0 ecotype. In a different study,
exposure of the WS ecotype to 300 ppb O3 for 6 hours showed a rapid induction of
genes leading to cell death, such as proteases, and downregulation or inactivation of
cell signaling genes, demonstrating an ineffective defense response in this O3
sensitive ecotype (Mahalingam et al., 2006).
The temporal response of plants to O3 exposure was evaluated in the Arabidopsis
Col-0 ecotype during a 6-hour exposure at 350 ppb O3 and for 6 hours after the
exposure was completed. Results of this study, shown in Figure 9-5 indicate that
genes associated with signal transduction and regulation of transcription were in the
class of early upregulated genes, while genes associated with redox homeostasis and
defense/stress response were in the class of late upregulated genes (Mahalingam et
al.. 2005).
A few studies have been conducted to evaluate transcriptome changes in response to
longer term chronic O3 exposures in woody plant species. Longer term exposures
resulted in the upregulation of genes associated with secondary metabolites,
including isoprenoids, polyamines and phenylpropanoids in 2-year-old seedlings of
the Mediterranean shrub Phillyrea latifolia exposed to 110 ppb O3 for 90 days
(Paolacci et al., 2007). In 3-year-old European beech saplings exposed to O3 for
20 months (with monthly average twice ambient O3 concentrations ranging from 11
to 80 ppb), O3-induced changes in gene transcription were similar to those observed
9-18
-------
for herbaceous species (Olbrich et al., 2009). Genes encoding proteins associated
with plant stress response, including ethylene biosynthesis, pathogenesis-related
proteins and enzymes detoxifying ROS, were upregulated. Some genes associated
with primary metabolism, cell structure, cell division and cell growth were reduced
(Olbrich et al., 2009). In a similar study using adult European beech trees, it was
determined that the magnitude of the transcriptional changes described above was far
greater in the saplings than in the adult trees exposed to the same O3 concentrations
for the same time period (Olbrich et al.. 2010).
The results from transcriptome studies described above have been substantiated by
results from proteome analysis in rice, poplar, European beech, wheat, and soybean.
Exposure of soybean to 120 ppb O3 for 12 hours/day for 3 days in growth chambers
resulted in decreases in the quantity of proteins associated with photosynthesis, while
proteins involved with antioxidant defense and carbon metabolism increased (Ahsan
et al., 2010). Young poplar plants exposed to 120 ppb O3 in a growth chamber for
35 days also showed significant changes in proteins involved in carbon metabolism
(Bohler et al., 2007). Declines in enzymes associated with carbon fixation, the Calvin
cycle and photosystem II were measured, while ascorbate peroxidase and enzymes
associated with glucose catabolism increased in abundance. In another study to
determine the impacts of O3 on both developing and fully expanded poplar leaves,
young poplars were exposed to 120 ppb O3 for 13 hours/day for up to 28 days
(Bohler et al., 2010). Impacts on protein quantity only occurred after the plants had
been exposed to O3 for 14 days, and at this point in time, several Calvin cycle
enzymes were reduced in quantity, while the effects on the light reactions appeared
later, at 21 days after beginning treatment. Some of the antioxidant enzymes
increased in abundance with O3 treatment, while others (ascorbate peroxidase) did
not. In relationship to leaf expansion, it was shown that O3 did not affect protein
quantity until leaves had reached full expansion, after about 7 days (Bohler et al..
2010).
Two-week-old rice seedlings exposed to varying levels of O3 (4, 40, 80, 120 ppb) in
a growth chamber for 9 days showed reductions in quantities of proteins associated
with photosynthesis and energy metabolism, and increases in some antioxidant and
defense related proteins (Feng et al., 2008a). A subsequent study of O3-treated rice
seedlings (exposed to 200 ppb O3 for 24 hours) focusing on the integration of
transcriptomics and proteomics, supported and further enhanced these results (Cho et
al., 2008). The authors found that of the 22,000 genes analyzed from the rice
genome, 1,535 were differentially regulated by O3. Those differentially regulated
genes were functionally categorized as transcription factors, MAPK cascades, those
encoding for enzymes involved in the synthesis of jasmonic acid (JA), ethylene (ET),
shikimate, tryptophan and lignin, and those involved in glycolysis, the citric acid
cycle, oxidative respiration and photosynthesis. The authors determined that the
proteome and metabolome (all small molecule metabolites in a cell) analysis
supported the results of the transcriptome changes described above (Cho et al.,
2008). This type of study, which ties together results from changes in gene
expression, protein quantity and activity, and metabolite levels, provides the most
9-19
-------
complete picture of the molecular and biochemical changes occurring in plants
exposed to a stressor such as O3.
Sarkar et al. (2010) compared proteomes of two cultivars of wheat grown in OTCs at
several O3 concentrations, including filtered air, ambient O3 (mean concentration
47 ppb), ambient +10 ppb and ambient + 20 ppb for 5 hours/day for 50 days.
Declines in the rate of photosynthesis and stomatal conductance were related to
decreases in proteins involved in carbon fixation and electron transport and increased
proteolysis of photosynthetic proteins such as the large subunit of ribulose-1,6-
bisphosphate carboxylase/oxygenase (Rubisco). Enzymes that take part in energy
metabolism, such as ATP synthesis, were also downregulated, while defense/stress
related proteins were upregulated in O3-treated plants. In comparing the two wheat
cultivars, Sarkar et al. (2010) found that while the qualitative changes in protein
expression between the two cultivars were similar, the magnitude of these changes
differed between the sensitive and tolerant wheat cultivars. Greater foliar injury and a
smaller decline in stomatal conductance was observed in the sensitive cultivar as
compared to the more tolerant cultivar, along with greater losses in photosynthetic
enzymes and higher quantities of antioxidant enzymes. Results from a three-year
exposure of European beech saplings to elevated O3 (AOT40 value was 52.6 ppm-h
for 2006, when trees were sampled) supported the results from the short-term
exposure studies described above (Kerner et al., 2011). The O3 treatment of the
saplings resulted in reductions in enzymes associated with the Calvin cycle, which
could lead to reduced carbon fixation. Enzymes associated with carbon
metabolism/catabolism were increased, and quantities of starch and sucrose were
reduced in response to the O3 treatment in these trees, indicating a potential impact
of O3 on overall carbon metabolism in long-term exposure conditions (Kerner et al.,
2011).
Transcriptome and proteome studies have provided valuable information about O3
effects on plants. These studies allow for simultaneous analysis of changes in the
expression patterns of many different genes and proteins, and also provide
information on how these molecules might interact with one another as a result of
plant exposure to oxidative stress. Gene and protein expression patterns generally
differ between O3-sensitive and tolerant plants, which could result from differential
uptake or detoxification of O3 or from differential regulation of the transcriptome
and proteome.
9-20
-------
(A)
Signaling
Transcription
Kedox hoincosiasis
Defense/stress response
PR proteins
(B)
HOST
12 hr
12 hr
Photosynthesis
Note: (A) Temporal profile of the oxidative stress response to O3. The biphasic O3-induced oxidative burst is represented in black,
with the ROS (reactive oxygen species) control measurements shown as a broken line. Average transcript profiles are shown for
early upregulated genes (yellow, peaks at 0.5-1 hours), and the 3 hours (blue), 4.5 hours (red) and 9-12 hours (green) late
upregulated genes and for the downregulated genes coding for photosynthesis proteins (brown). PR = pathogenesis related. (B)
Diagrammatic representation of redox regulation of the oxidative stress response. TF = transcription factor; SA = salicylic acid.
Source: Reprinted with permission of Springer (Mahalingam et al.. 2005).
Figure 9-5 Composite diagram of major themes in the temporal evolution of
the genetic response to Os stress.
All of these studies describe common trends for changes in gene and protein
expression which occur in a variety of plant species exposed to O3. While genes
associated with carbon assimilation and general metabolism are typically
downregulated, genes associated with signaling, catabolism, and defense are
upregulated. The magnitude of these changes in gene and protein expression appears
to be related to plant species, age and their sensitivity or tolerance to O3.
9-21
-------
9.3.3.3 Role of Phytohormones in Plant Response to Ozone
Many studies of O3 effects on plants have analyzed the importance of plant
hormones such as SA, ET and JA in determining plant response to O3. The 2006 O3
AQCD documents the O3-induced production of ET and its role in promoting the
formation of leaf lesions. Transcriptome analysis and the use of a variety of mutants
have allowed for further elucidation of the complex interactions between SA, ET, JA
and the role of abscisic acid (ABA) in mediating plant response to O3 (Ludwikow
and Sadowski, 2008). In addition to their roles in signaling pathways, phytohormones
also appear to regulate, and be regulated by, the MAPK signaling cascades described
previously. Most evidence suggests that while ET and SA are needed to develop
O3-induced leaf lesions, JA acts antagonistically to SA and ET to limit the lesions
(Figure 9-6) (Kangasjarvi et al., 2005).
The rapid production of ET in O3 treated plants has been described in many plant
species and has been further characterized through the use of a variety of mutants
that either over-produce or are insensitive to ET. Production of stress ET in
O3-treated plants, which is thought to be part of a wounding response, was found to
be correlated to the degree of injury development in leaves (U.S. EPA. 2006b). More
recent studies have supported these conclusions and have also focused on the
interactions occurring between several oxidative-stress induced phytohormones.
Yoshida et al. (2009) determined that ET likely amplifies the oxidative signal
generated by ROS, thereby promoting lesion formation. By analyzing the O3-induced
transcriptome of several Arabidopsis mutants of the Col-0 ecotype, Tamaoki et al.
(2003) determined that at 12 hours after initiating the O3 exposure (200 ppb for
12 hours), the ET and JA signaling pathways were the main pathways used to
activate plant defense responses, with a lesser role for SA. The authors also
demonstrated that low levels of ET production could stimulate the expression of
defense genes, rather than promoting cell death which occurs when ET production is
high. Tosti et al. (2006) supported these findings by showing that plant exposure to
O3 not only results in activation of the biosynthetic pathways of ET, JA and SA, but
also increases the expression of genes related to the signal transduction pathways of
these phytohormones in O3-treated Arabidopsis plants (300 ppb O3 for 6 hours).
Conversely, in the O3 sensitive Ws ecotype, its sensitivity may, in part, be due to
intrinsically high ET levels leading to SA accumulation, and the high ET and SA
may act to repress JA-associated genes, which would serve to inhibit the spread of
lesions (Mahalingam et al., 2006). Ogawa et al. (2005) found that increases in SA in
O3-treated plants leads to the formation of leaf lesions in tobacco plants exposed to
200 ppb O3 for 6 hours. Furthermore, in transgenic tobacco plants with reduced
levels of ET production in response to O3 exposure, several genes encoding for
enzymes in the biosynthetic pathway of SA were suppressed, suggesting that SA
levels are, in part, controlled by ET in the presence of O3.
Exposure of the Arabidopsis mutant rcdl to acute doses of O3 (250 ppb O3 for
8 hours/day for 3 days) resulted in programmed cell death (PCD) and the formation
of leaf lesions (Overmyer et al., 2000). They determined that the observed induction
of ET synthesis promotes cell death, and that ET perception and signaling are
9-22
-------
required for the accumulation of superoxide, which leads to cell death and
propagation of lesions. Jasmonic acid, conversely, contains the spread of leaf lesions
(Overmyer et al, 2000). Transcriptome analysis of several Arabidopsis mutants,
which are insensitive to SA, ET and JA, exposed to 12-h of 200 ppb O3 showed that
approximately 78 of the upregulated genes measured in this study were controlled by
ET and JA signaling pathways, while SA signaling pathways were suggested to
antagonize ET and JA pathways (Tamaoki et al.. 2003). In a subsequent
transcriptome study on the Col-0 ecotype exposed to 150 ppb O3 for 48-h, JA and ET
synthesis were downregulated, while SA was upregulated in O3-treated plants.
In cotton plants exposed to a range of O3 concentrations (0-120 ppb) and methyl
jasmonate (MeJA), Grantz et al. (201 Ob) determined that exogenous applications of
MeJA did not protect plants from chronic O3 exposure.
Abscisic acid has been investigated for its role in regulating stomatal aperture and
also for its contribution to signaling pathways in the plant. The role of ABA and the
interaction between ABA and H2O2 in O3-induced stomatal closure was described in
the 2006 O3 AQCD. It was determined that the presence of H2O2, which is formed
from O3 degradation, increases the sensitivity of guard cells to ABA and, therefore,
more readily results in stomatal closure. More recently, it was determined that
synthesis of ABA was induced in O3-treated Arabidopsis plants (250-350 ppb O3 for
6 hours), with a more pronounced induction in the O3 sensitive rcd3 mutant as
compared to the wildtype Col-0 (Overmyer et al., 2008). The rcd3 mutant also
exhibited a lack of O3-induced stomatal closure, and the RCD3 protein has been
shown to be required for slow anion channels (Overmyer et al., 2008). Ludwikow et
al. (2009) used Arabidopsis ABIltd mutants, in which a key negative regulator of
ABA action (abscisic acid insensitivel protein phosphatase 2C) has been knocked
out, to examine O3 responsive genes in this mutant compared to the Arabidopsis
Col-0. Results of this study indicate a role for ABU in negatively regulating the
synthesis of both ABA and ET in O3-treated plants (350 ppb O3 for 9 hours).
Additionally, ABU may stimulate JA-related gene expression, providing evidence
for an antagonistic interaction between ABA and JA signaling pathways (Ludwikow
et al.. 2009).
Nitric oxide (NO) has also been shown to play a role in regulating gene expression in
plants in response to O3 exposure. However, little is known to date about NO and its
role in the complex interactions of molecules in response to O3. Exposure of tobacco
to O3 (150 ppb for 5 hours) stimulated NO and NO-dependent ET production, while
NO production itself did not depend on the presence of ET (Ederli et al., 2006).
Analysis of O3-treated Arabidopsis indicated the possibility of a dual role for NO in
the initiation of cell death and later lesion containment (Ahlfors et al., 2009).
While much work remains to be done to better elucidate what determines plant
sensitivity and response to O3, it is clear that the mechanism for response to O3 and
signal transduction is very complex. Many of the phytohormones and other signaling
molecules thought to be involved in these processes are interactive and depend upon
a variety of other factors, which could be either internal or external to the plant. This
results in a highly dynamic and complex system, capable of resulting in a spectrum
9-23
-------
of plant sensitivity to oxidative stress and generating a variety of plant responses to
that stress.
Cell
death
Note: Ozone-derived radicals induce endogenous ROS production (1) which results in salicylic acid (SA) accumulation and
programmed cell death; (2) Cell death triggers ethylene (ET) production, which is required for the continuing ROS production
responsible for the propagation of cell death; (3) Jasmonates counteract the progression of the cycle by antagonizing the cell
death promoting function of SA and ET; (4) Abscisic acid (ABA) antagonizes ET function in many situations and might also have
this role in O3-induced cell death; (5) Mutually antagonistic interactions between ET, SA and jasmonic acid (JA) are indicated with
red bars.
Source: Reprinted with permission of Blackwell Publishing Ltd. (Kangasiarvi et al.. 2005).
Figure 9-6 The oxidative cell death cycle.
9.3.4 Detoxification
9.3.4.1 Overview of Ozone-induced Defense Mechanisms
Plants are exposed to an oxidizing environment on a continual basis, and many
reactions that are part of the basic metabolic processes, such as photosynthesis and
respiration, generate ROS. As a result, there is an extensive and complex mechanism
in place to detoxify these oxidizing radicals, including both enzymes and
metabolites, which are located in several locations in the cell and also in the apoplast
of the cell. As O3 enters the leaf through open stomata, the first point of contact of
9-24
-------
O3 with the plant is likely in the apoplast, where it breaks down to form oxidizing
radicals such as H2O2, O2~, HO- and HO2. Another source of oxidizing radicals is an
oxidative burst, generated by a membrane-bound NADPH oxidase enzyme, which is
recognized as an integral component of the plant's defense system against pathogens
(Schraudner et al., 1998). Antioxidant metabolites and enzymes located in the
apoplast are thought to form a first line of defense by detoxifying O3 and/or the ROS
that are formed as breakdown products of O3 (Section 9.3.2). However, even with the
presence of several antioxidants, including ascorbate, the redox buffering capacity of
the apoplast is far less than that of the cytoplasm, as it lacks the regeneration systems
necessary to retain a reduced pool of antioxidants (Foyer and Noctor. 2005b).
Redox homeostasis is regulated by the presence of a pool of antioxidants, which are
typically found in a reduced state and detoxify ROS produced by oxidases or electron
transport components. As ROS increase due to environmental stress such as O3, it is
unclear whether the antioxidant pool can maintain its reduced state (Foyer and
Noctor, 2005b). As such, not only the quantity and types of antioxidant enzymes and
metabolites present, but also the cellular ability to regenerate those antioxidants are
important considerations in mechanisms of plant tolerance to oxidative stress
(Dizengremel et al., 2008). Molecules such as glutathione (GSH), thioredoxins and
NADPH play very important roles in this regeneration process; additionally, it has
been hypothesized that alterations in carbon metabolism would be necessary to
supply the needed reducing power for antioxidant regeneration (Dizengremel et al.,
2008).
9.3.4.2 Role of Antioxidants in Plant Defense Responses
Ascorbate has been the focus of many different studies as an antioxidant metabolite
that protects plants from exposure to O3. It is found in several cellular locations,
including the chloroplast, the cytosol and the apoplast (Noctor and Foyer, 1998).
Ascorbate is synthesized in the cell and transported to the apoplast. Apoplastic
ascorbate can be oxidized to dehydroascorbate (DHA) with exposure to O3 and is
then transported back to the cytoplasm. Here, DHA is reduced to ascorbate by the
enzyme dehydroascorbate reductase (DHAR) and reduced GSH, which is part of the
ascorbate-glutathione cycle (Noctor and Foyer, 1998). Many studies have focused on
evaluating whether ascorbate is the primary determining factor in differential
sensitivity of plants to O3. An evaluation of several species of wildflowers in Great
Smoky Mountains National Park showed a correlation between higher quantities of
reduced apoplastic ascorbate and lower levels of foliar injury from O3 exposure in a
field study on tall milkweed plants (Asclepsias exaltata L.) (Burkey et al., 2006;
Souzaet al., 2006). Cheng et al. (2007) exposed two soybean cultivars to elevated O3
(77 ppb) and filtered air for 7 hours/day for 6 days. The differences in sensitivity
between the two cultivars could not be explained by differential O3 uptake or by the
fraction of reduced ascorbate present in the apoplast. However, total antioxidant
capacity of the apoplast was 2-fold higher in the tolerant Essex cultivar as compared
to the sensitive Forrest cultivar, indicating that there may be other compounds in the
9-25
-------
leaf apoplast that scavenge ROS. D'Haese et al. (2005) exposed the NC-S (sensitive)
and NC-R (resistant) clones of white clover (Trifolium repens) to 60 ppb O3 for
7 hours/day for 5 days in environmental chambers. Surprisingly, the NC-S clone had
a higher constitutive concentration of apoplastic ascorbate with a higher redox status
than the NC-R clone. However, the redox status of symplastic GSH was higher in
NC-R, even though the concentration of GSH was not higher than in NC-S.
In addition, total symplastic antioxidative capacity was not a determining factor in
differential sensitivity between these two clones. Severino et al. (2007) also
examined the role of antioxidants in the differential sensitivity of the two white
clover clones by growing them in the field for a growing season and then exposing
them to elevated O3 (100 ppb for 8 hours/day for 10 days) in OTC at the end of the
field season. The NC-R clone had greater quantities of total ascorbate and total
antioxidants than the NC-S clone at the end of the experiment. In snap bean, plants of
the O3 tolerant Provider cultivar had greater total ascorbate and more ascorbate in the
apoplast than the sensitive SI56 cultivar after exposure to 71 ppb O3 for 10 days in
OTC (Burkey et al.. 2003). While most of the apoplastic ascorbate was in the
oxidized form, the ratio of reduced ascorbate to total ascorbate was higher in
Provider than SI56, indicating that Provider is better able to maintain this ratio to
maximize plant protection from oxidative stress. Exposure of two wheat varieties to
ambient (7-h average 44 ppb O3) and elevated (7-h average 56 ppb O3) O3 for
60 days in open-air field conditions showed higher concentrations of reduced
ascorbate in the apoplast in the tolerant Y16 variety than the more sensitive Y2
variety, however no varietal differences were seen in the decrease in reduced
ascorbate quantity in response to O3 exposure (Feng et al.. 2010). To evaluate
whether O3 affected apoplastic concentrations of ascorbic acid and phenolic
compounds, wildtype Arabidopsis thaliana (Col-0, Ler-0) and null mutants lacking
sinapoyl and flavonol glycosides were exposed to either 125 or 175 ppb O3 for up to
2 days. The authors determined that ascorbic acid, which was found in very low
quantities in the reduced form, and the phenolic compounds did not play an
important role in protecting plants from O3 injury (Booker et al., 2012). While there
is much evidence that supports an important role for ascorbate, particularly
apoplastic ascorbate, in protecting plants from oxi dative stressors such as O3, it is
also clear that there is much variation in the importance of ascorbate for different
plant species and differing exposure conditions. Additionally, the work of several
authors suggests that there may be other compounds in the apoplast which have the
capacity to act as antioxidants.
While the quantities of antioxidant metabolites such as ascorbate are an important
indicator of plant tolerance to O3, the ability of the plant to recycle oxidized
ascorbate efficiently also plays a large role in determining the plant's ability to
effectively protect itself from sustained exposure to oxi dative stress. Tobacco plants
over-expressing DHAR were better protected from exposure to either chronic
(100 ppb O3 4 hours/day for 30 days) or acute (200 ppb O3 for 2 hours) O3
conditions than control plants and those with reduced expression of DHAR (Chen
and Gallic. 2005). The DHAR over-expressing plants exhibited an increase in guard
cell ascorbic acid, leading to a decrease in stomatal responsiveness to O3 and an
increase in stomatal conductance and O3 uptake. Despite this, the presence of higher
9-26
-------
levels of ascorbic acid led to a lower oxidative load and a higher level of
photo synthetic activity in the DHAR over-expressing plants (Chen and Gallic, 2005).
A subsequent study with tobacco plants over-expressing DHAR confirmed some of
these results. Levels of ascorbic acid were higher in the transgenic tobacco plants,
and they exhibited greater tolerance to O3 exposure (200 ppb O3) as demonstrated by
higher photosynthetic rates in the transgenic plants as compared to the control plants
(Eltaveb et al.. 2006). Over-expression of monodehydroascorbate reductase (MDAR)
in tobacco plants also showed enhanced stress tolerance in response to O3 exposure
(200 ppb O3), with higher rates of photosynthesis and higher levels of reduced
ascorbic acid as compared to controls (Eltaveb et al.. 2007). Results of these studies
demonstrate the importance of ascorbic acid as a detoxification mechanism in some
plant species, and also emphasize that the recycling of oxidized ascorbate to maintain
a reduced pool of ascorbate is a factor in determining plant tolerance to oxidative
stress.
The roles of other antioxidant metabolites and enzymes, including GSH, catalase
(CAT), peroxidase (POD) and superoxide dismutase (SOD), were comprehensively
reviewed in the 2006 O3 AQCD. Based on the review of the literature, no conclusive
and consistent effects of O3 on the quantity of GSH and CAT could be identified.
Both apoplastic and cytosolic POD activity increased in response to O3 exposure,
while various isoforms of SOD showed inconsistent changes in quantity in response
to O3. Additional studies have been conducted to further elucidate the roles of these
antioxidant enzymes and metabolites in protecting plants from oxidative stress.
Superoxide dismutase and POD activities were measured in both the tolerant Bel B
and sensitive Bel W3 tobacco cultivars exposed to ambient O3 concentrations for
2 weeks 3 times throughout a growing season (Borowiak et al., 2009). In this study,
SOD and POD activity, including that of several different isoforms, increased in both
the sensitive and tolerant tobacco cultivars with exposure to O3, however the
isoenzyme composition for POD differed between the sensitive and tolerant tobacco
cultivars (Borowiak et al.. 2009) Tulip poplar (Liriodendron tulipiferd) trees exposed
to increasing O3 concentrations (from 100 to 300 ppb O3 during a 2-week period)
showed increases in activities of SOD, ascorbate peroxidase (APX), glutathione
reductase (GR), MDAR, DHAR, CAT and POD in the 2-week period, although
individual enzyme activities increased at different times during the 2-week period
(Rvang et al.. 2009).
Longer, chronic O3 exposures in trees revealed increases in SOD and APX activity in
Quercus mongolica after 45 days of plant exposure to 80 ppb O3, which were
followed by declines in the activities and quantities of these enzymes after 75 days of
exposure (Yan et al.. 2010). Similarly, activities of SOD, APX, DHAR, MDAR, and
GR increased in Gingko biloba trees during the first 50 days of exposure to 80 ppb
O3, followed by decreases in activity below control values after 50 days of exposure
(He et al., 2006). Soybean plants exposed to 70 or 100 ppb O3 for 4 hours/day over
the course of a growing season showed elevated POD activity and a decrease in CAT
activity at 40 and 60 days after germination (Singh et al., 2010a).
9-27
-------
Antioxidant enzymes and metabolites have been shown to play an important role in
determining plant tolerance to O3 and mediating plant responses to O3. However,
there is also some evidence to suggest that the direct reaction of ascorbate with O3
could lead to the formation of secondary toxicants, such as peroxy compounds,
which may act upon signal transduction pathways and modulate plant response to O3
(Sandermann. 2008). Therefore, the role of ascorbate and other antioxidants and their
interaction with other plant responses to O3, such as the activation of signal
transduction pathways, is likely far more complex than is currently understood.
9.3.5 Effects on Primary and Secondary Metabolism
9.3.5.1 Light and Dark Reactions of Photosynthesis
Declines in the rate of photosynthesis in O3-treated plants have been documented for
many different plant species (Booker et al.. 2009: Wittig et al.. 2007: U.S. EPA.
2006b). The 2006 O3 AQCD described the mechanism by which plant exposure to
O3 reduces carboxylation capacity, and the more recent scientific literature confirms
these findings. While several measures of the light reactions of photosynthesis are
sensitive to exposure to O3 (see below), photosynthetic carbon assimilation is
generally considered to be more affected by pollutant exposure, resulting in an
overall decline in photosynthesis (Guidi and Degl'lnnocenti. 2008: Heath. 2008:
Fiscus et al.. 2005). Loss of carbon assimilation capacity has been shown to result
from declines in the quantity and activity of Rubisco (Calatavud et al.. 2010:
Goumenakietal.. 2010: Singh et al.. 2009: Bagard et al.. 2008: Calatavud et al..
2007a: Crous et al.. 2006). Experimental evidence suggests that both decreases in
Rubisco synthesis and enhanced degradation of the protein contribute to the
measured reduction in its quantity. Additionally, the reduction in Rubisco quantity
has been associated with the O3-induced oxidative modification of the enzyme,
which is evidenced by the increases in carbonyl groups on the protein after plant
exposure to O3 (U.S. EPA. 2006b). Reduced carbon assimilation has been linked to
reductions in biomass and yield (Wang et al.. 2009b: He et al.. 2007: Novak et al..
2007: Gregg et al.. 2006: Keutgen et al.. 2005). Recent studies evaluating O3 induced
changes in the transcriptome and proteome of several different species confirm these
findings. Levels of mRNA for rbcS (the gene that encodes the small subunit [SSU] of
the RuBisCO protein [ribulose-l,5-bisphosphate carboxylase/oxygenase, a major
stromal enzyme involved in carbon fixation by plants]) declined in European beech
saplings exposed to 300 ppb O3 for 8 hours/day for up to 26 days (Olbrich et al..
2005). Similar declines in rbcS mRNA were also measured in the beech saplings in a
free air exposure system over a course of two growing seasons (Olbrich et al.. 2009).
Proteomics studies have also confirmed the effects of O3 on proteins involved in
carbon assimilation. Reductions in quantities of the small and large subunit (rbcL) of
Rubisco and Rubisco activase were measured in soybean plants exposed to 120 ppb
O3 for 3 days in growth chambers (Ahsan et al.. 2010). Exposure of young poplar
trees to 120 ppb O3 for 35 days in exposure chambers resulted in reductions of
9-28
-------
Rubisco, Rubisco activase, and up to 24 isoforms of Calvin cycle enzymes, most of
which play a role in regenerating the CO2 acceptor molecule, ribulose-1,5-
bisphosphate (Bohler et al., 2007). Reductions in protein quantity of both the small
and large subunit of Rubisco were seen in wheat plants exposed to ambient (average
concentration 47.3 ppb O3) and elevated O3 (ambient + 10 or 20 ppb O3) in open-top
chambers for 5 hours/day for 50 days (Sarkar et al.. 2010). Lettuce plants exposed to
100 ppb O3 in growth chambers for 8 hours/day for 3 weeks also showed reductions
in transcript and protein levels of the small and large subunits of Rubisco and
Rubisco activase (Goumenaki et al.. 2010).
Reductions in photosynthesis are not only related to declines in the quantity of
Rubisco, but also of its activity level. The maximum carboxylation rate (Vcmax) has
been shown to decline in plants species exposed to O3, including lettuce (Goumenaki
et al., 2010), white clover (Crous et al., 2006), young poplar trees (Bagard et al.,
2008) and evergreen deciduous shrubs (Calatavud et al., 2010). While a significant
proportion of the reduction in Vcmax is caused by declines in the quantity of Rubisco,
other contributors to changes in Vcmax result from reductions in the quantity and
activity of Rubisco activase, an enzyme which prepares Rubisco for carbamylation
by accelerating the release of bound sugar phosphates. Reductions in Rubisco
activase quantity have been observed in several studies evaluating the effects of O3
on the proteomes of poplar (Bohler et al., 2007), European beech (Kerner et al.,
2011) and soybean (Ahsan et al., 2010).
In addition to impacts on carbon assimilation, the deleterious effects of O3 on the
photosynthetic light reactions have received more attention in recent years.
Chlorophyll fluorescence provides a useful measure of changes to the photosynthetic
process from exposure to oxidative stress. Decreases in the Fv/Fm ratio (a measure of
the maximum efficiency of Photosystem II) in dark adapted leaves indicate a decline
in the efficiency of the PSII photosystems and a concomitant increase in non-
photochemical quenching (Guidi and Degl'lnnocenti, 2008; Scebba et al., 2006).
Changes in these parameters have been correlated to differential sensitivity of plants
to the pollutant. In a study to evaluate the response of 4 maple species to O3 (exposed
to an 8-h avg of 51 ppb for ambient and 79 ppb for elevated treatment in OTC), the 2
species which were most sensitive based on visible injury and declines in CO2
assimilation also showed the greatest decreases in Fv/Fm in symptomatic leaves.
In asymptomatic leaves, CO2 assimilation decreased significantly but there was no
significant decline in Fv/Fm (Calatavud et al., 2007a). Degl'lnnocenti et al. (2007)
measured significant decreases in Fv/Fm in young and symptomatic leaves of a
resistant tomato genotype (line 93.1033/1) in response to O3 exposure (150 ppb O3
for 3 hours in a growth chamber), but only minor decreases in asymptomatic leaves
with no associated changes in net photosynthetic rate. In the O3 sensitive tomato
cultivar Cuor Di Bue, the Fv/Fm ratio did not change, while the photosynthetic rate
declined significantly in asymptomatic leaves (Degl'lnnocenti et al., 2007). In two
soybean cultivars, Fv/Fm also declined significantly with plant exposure to O3
(Singh et al., 2009). It appears that in asymptomatic leaves, photoinhibition, as
indicated by a decrease in Fv/Fm, is not the main reason for a decline in
photosynthesis.
9-29
-------
An evaluation of photosynthetic parameters of two white clover (Trifolium repens cv.
Regal) clones that differ in their O3 sensitivity revealed that O3 (40-110 ppb O3 for
7 hours/day for 5 days) increased the coefficient of non-photochemical quenching
(qisip) in both the resistant (NC-R) and sensitive (NC-S) clones, however q^ was
significantly lower for the sensitive clone (Crous et al., 2006). Sensitive Acer clones
had a lower coefficient of non-photochemical quenching, while exposure to O3
increased qNp in both sensitive and tolerant clones (Calatavud et al.. 2007a). While
exposure to O3 also increased qM> in tomato, there were no differences in the
coefficient of photochemical quenching between cultivars thought to be differentially
sensitive to O3 (Degl'Innocenti et al.. 2007). Higher q^ as a result of exposure to O3
indicates a reduction in the proportion of absorbed light energy being used to drive
photochemistry. A lower coefficient of non-photochemical quenching in O3 sensitive
plants could indicate increased vulnerability to ROS generated during exposure to
oxidative stress (Crous et al.. 2006).
Most of the research on O3 effects on photosynthesis has focused on C3 (Calvin
cycle) plants because C4 (Hatch-Slack) plants have lower stomatal conductance and
are, therefore, thought to be less sensitive to O3 stress. However, some studies have
been conducted to evaluate the effects of O3 on C4 photosynthesis. In older maize
leaves, Leitao et al. (2007c; 2007a) found that the activity, quantity and transcript
levels of both Rubisco and phosphoenolpyruvate carboxylase (PEPc) decreased as a
function of rising O3 concentration. In younger maize leaves, the quantity, activity,
and transcript levels of the carboxylases were either increased or unaffected in plants
exposed to 40 ppb O3 for 7 hours/day for 28-33 days, but decreased at 80 ppb (Leitao
et al., 2007b; Leitao et al., 2007c). In another study, Grantz et al. (2009) reported that
O3 exposures (4, 58, and 114 ppb, 12-h mean) decreased sugarcane biomass
production by more than one third and allocation to roots by more than two thirds.
9.3.5.2 Respiration and Dark Respiration
While much research emphasis regarding O3 effects on plants has focused on the
negative impacts on carbon assimilation, other studies have measured impacts on
catabolic pathways such as shoot respiration and photorespiration. Generally, shoot
respiration has been found to increase in plants exposed to O3. Bean plants exposed
to ambient (average 12-h mean 43 ppb) and twice ambient (average 12-h mean
80 ppb) O3 showed increases in respiration. When mathematically partitioned, the
maintenance coefficient of respiration was significantly increased in O3 treated
plants, while the growth coefficient of respiration was not affected (Amthor, 1988).
Loblolly pines were exposed to ambient (12-h daily mean was 45 ppb) and twice
ambient (12-h daily mean was 86 ppb) O3 for 12 hours/day for approximately
seven months per year for 3 and 4 years. While photosynthetic activity declined with
the age of the needles and increasing O3 concentration, enzymes associated with
respiration showed higher levels of activity with increasing O3 concentration
(Dizengremel et al., 1994). In their review on the role of metabolic changes in plant
redox status after O3 exposure, Dizengremel et al. (2009) summarized multiple
9-30
-------
studies in which several different tree species were exposed to O3 concentrations
ranging from ambient to 200 ppb O3 for at least several weeks. In all cases, the
activity of enzymes, including phosphofructokinase, pyruvate kinase and fumarase,
which are part of several catabolic pathways, were increased in O3 treated plants.
Photorespiration is a light-stimulated process which consumes O2 and releases CO2.
While it has been regarded as a wasteful process, more recent evidence suggests that
it may play a role in photoprotection during photosynthesis (Bagard et al., 2008).
The few studies that have been conducted on O3 effects on photorespiration suggest
that rates of photorespiration decline concomitantly with rates of photosynthesis.
Soybean plants were exposed to ambient (daily averages 43-58 ppb) and 1.5 ambient
O3 (daily averages 63-83 ppb) O3 in OTCs for 12 hours/day for 4 months. Rates of
photosynthesis and photorespiration and photorespiratory enzyme activity declined
only at the end of the growing season and did not appear to be very sensitive to O3
exposure (Booker et al., 1997). Young hybrid poplars exposed to 120 ppb O3 for
13 hours/day for 35 days in phytotron chambers showed that effects on
photorespiration and photosynthesis were dependent upon the developmental stage of
the leaf. While young leaves were not impacted, reductions in photosynthesis and
photorespiration were measured in fully expanded leaves (Bagard et al., 2008).
9.3.5.3 Secondary Metabolism
Transcriptome analysis of Arabidopsis plants has revealed modulation of several
genes involved in plant secondary metabolism (Ludwikow and Sadowski. 2008).
Phenylalanine ammonia lyase (PAL) has been the focus of many studies involving
plant exposure to O3 due to its importance in linking the phenylpropanoid pathway of
plant secondary metabolism to primary metabolism in the form of the shikimate
pathway. Genes encoding several enzymes of the phenylpropanoid pathway and
lignin biosynthesis were upregulated in transcriptome analysis of Arabidopsis plants
(Col-0) exposed to 350 ppb O3 for 6 hours, while 2 genes involved in flavonoid
biosynthesis were downregulated (Ludwikow et al., 2004). Exposure of Arabidopsis
(Col-0) to lower O3 concentrations (150 ppb for 8 hours/day for 2 days) resulted in
the induction of 11 transcripts involved in flavonoid synthesis. In their exposure of
2-year-old Mediterranean shrub Phillyrea latifolia to 110 ppb O3 for 90 days,
Paolacci et al. (2007) identified four clones that were upregulated and corresponded
to genes involved in the synthesis of secondary metabolites, such as isoprenoids,
polyamines and phenylpropanoids. Upregulation of genes involved in isoprene
synthesis was also observed \nMedicago trunculata exposed to 300 ppb O3 for
6 hours, while genes encoding enzymes of the flavonoid synthesis pathway were
either upregulated or downregulated (Puckette et al., 2008). Exposure of red clover to
1.5 x ambient O3 (average concentrations of 32.4 ppb) for up to 9 weeks in an open
field exposure system resulted in increases in leaf total phenolic content. However,
the types of phenolics that were increased in response to O3 exposure differed
depending upon the developmental stage of the plant. While almost all of the 31
different phenolic compounds measured increased in quantity initially during the
9-31
-------
exposure, after 3 weeks the quantity of isoflavones decreased while other phenolics
increased (Saviranta et al.. 2010). Exposure of beech saplings to ambient and
2 x ambient O3 concentrations over 2 growing seasons resulted in the induction of
several enzymes which contribute to lignin formation, while enzymes involved in
flavonoid biosynthesis were downregulated (Olbrich et al., 2009). Exposure of
tobacco Bel W3 to 160 ppb O3 for 5 hours showed upregulation of almost all genes
encoding for enzymes which are part of the prechorismate pathway (Janzik et al..
2005). Isoprenoids can serve as antioxidant compounds in plants exposed to
oxidative stress (Paolacci et al.. 2007).
The prechorismate pathway is the pathway leading to the formation of chorismate, a
precursor to the formation of the aromatic amino acids tryptophan, tyrosine and
phenylalanine. These amino acids are precursors for the formation of many
secondary aromatic compounds, and, therefore, the prechorismate pathway
represents a branch-point in the regulation of metabolites into either primary or
secondary metabolism (Janzik et al., 2005). Exposure of the O3 sensitive Bel W3
tobacco cultivar at 160 ppb for 5 hours showed an increase in transcript levels of
most of the genes encoding enzymes of the prechorismate pathway. However,
shikimate kinase (SK) did not show any change in transcript levels and only one of
three isoforms of DAHPS (3-deoxy-D-arabino-heptulosonat-7-phosphate synthase),
the first enzyme in this pathway, was induced by O3 exposure (Janzik et al., 2005).
Differential induction of DAHPS isoforms was also observed in European beech
after 40 days of exposure to 150-190 ppb O3. At this time point in the beech
experiment, transcript levels of shikimate pathway enzymes, including SK, were
generally strongly induced after an only weak initial induction after the first 40 days
of exposure. Both soluble and cell-wall bound phenolic metabolites showed only
minimal increases in response to O3 for the duration of the exposure period (Alonso
et al.. 2007). Total leaf phenolics decreased with leafage in Populus nigra exposed
to 80 ppb O3 for 12 hours/day for 14 days. Ozone increased the concentration of total
leaf phenolics in newly expanded leaves, with the greatest increases occurring in
compounds such as quercitin glycoside, which has a high antioxidant capacity (Fares
et al.. 201 Ob). While several phenylpropanoid pathway enzymes were induced in two
poplar clones exposed to 60 ppb O3 for 5 hours/day for 15 days, the degree of
induction differed between the two clones. In the tolerant 1-214 clone, PAL activity
increased 9-fold in O3-treated plants as compared to controls, while there was no
significant difference in PAL activity in the sensitive Eridano clone (Di Baccio et al..
2008).
Polyamines such as putrescine, spermidine and spermine play a variety of roles in
plants and have been implicated in plant defense responses to both abiotic and biotic
stresses. They exist in both a free form and conjugated to hydroxycinnamic acids.
Investigations on the role of polyamines have found that levels of putrescine increase
in response to oxi dative stress. This increase stems largely from the increase in the
activity of arginine decarboxylase (ADC), a key enzyme in the synthesis of
putrescine (Groppa and Benavides. 2008). Langebartels et al. (1991) described
differences in putrescine accumulation in O3-treated tobacco plants exposed to
several O3 concentrations, ranging from 0-400 ppb for 5-7 hours. A large and rapid
9-32
-------
increase in putrescine occurred in the tolerant Bel B cultivar and only a small
increase in the sensitive Bel W3 cultivar, which occurred only after the formation of
necrotic leaf lesions. Van Buuren et al. (2002) further examined the role of
polyamines in these two tobacco cultivars during an acute (130 ppb O3 for 7-h in a
growth chamber) exposure. They found that while free putrescine accumulated in
undamaged tissue of both cultivars, conjugated putrescine predominantly
accumulated in tissues undergoing cell death after plant exposure to O3 (van Buuren
et al.. 2002). The authors suggest that while free putrescine may not play a role in
conferring tolerance in the Bel B cultivar, conjugated putrescine may play a role in
O3-induced programmed cell death in Bel W3 plants.
Isoprene is emitted by some plant species and represents the predominant biogenic
source of hydrocarbon emissions in the atmosphere (Guenther et al., 2006). In the
atmosphere, the oxidation of isoprene by hydroxyl radicals can enhance O3
formation in the presence of NOX, thereby impacting the O3 concentration that plants
are exposed to. While isoprene emission varies widely between species, it has been
proposed to stabilize membranes and provide those plant species that produce it with
a mechanism of thermotolerance (Sharkey et al., 2008). It has also been suggested
that isoprene may act as an antioxidant compound to scavenge O3 (Loreto and
Velikova, 2001). Recent studies using a variety of plant species have shown
conflicting results in trying to understand the effects of O3 on isoprene emission.
Exposure to acute doses of O3 (300 ppb for 3-h) in detached leaves ofPhragmites
australis resulted in stimulation of isoprene emissions (Velikova et al., 2005).
Similar increases in isoprene emissions were measured in Populus nigra after
exposure to 100 ppb O3 for 5 days continuously (Fares et al., 2008). Isoprene
emission in attached leaves of Populus alba, which were exposed to 150 ppb O3 for
11 hours/day for 30 days inside cuvettes, was inhibited, while isoprene emission and
transcript levels of isoprene synthase mRNA were increased in the leaves exposed to
ambient O3 (40 ppb), which were located above the leaves enclosed in the exposure
cuvettes (Fares et al.. 2006). Exposure of 2 genotypes of hybrid poplar to 120 ppb O3
for 6 hours/day for 8 days resulted in a significant reduction in isoprene emission in
the O3-sensitive but not the tolerant genotype (Ryan et al.. 2009). Similarly, O3
treatment (80 ppb 12 hours/day for 14 days) of Populus nigra showed that isoprene
emission was reduced in the treated plants relative to the control plants (Fares et al..
201 Ob). Based on results of this and other studies, Fares et al. (201 Ob) concluded that
the isoprenoid pathway may be induced in plants exposed to acute O3 doses, while at
lower doses isoprene emission may be inhibited. Vickers et al. (2009) developed
transgenic tobacco plants with the isoprene synthase gene from Populus alba and
exposed them to 120 ppb O3 for 6 hours/day for 2 days. They determined that the
wildtype plants showed significantly more O3 damage, including the development of
leaf lesions and a decline in photosynthetic rates, than the transgenic,
isoprene-emitting plants. Transgenic plants also accumulated less H2O2 and had
lower levels of lipid peroxidation following exposure to O3 than the wildtype plants
(Vickers et al.. 2009). These results indicate that isoprene may have a protective role
for plants exposed to oxidative stress.
9-33
-------
9.3.6 Summary
The results of recent studies on the effects of O3 stress on plants support and
strengthen those reported in the 2006 O3 AQCD. The most significant new body of
evidence since the 2006 O3 AQCD comes from research on molecular mechanisms
of the biochemical and physiological changes observed in many plant species in
response to O3 exposure. Recent studies have employed new techniques, such as
those used in evaluating transcriptomes and proteomes to perform very
comprehensive analyses of changes in gene transcription and protein expression in
plants exposed to O3. These newer molecular studies not only provide very important
information regarding the many mechanisms of plant responses to O3, they also
allow for the analysis of interactions between various biochemical pathways which
are induced in response to O3. However, many of these studies have been conducted
in artificial conditions with model plants, which are typically exposed to very high,
short doses of O3. Therefore, additional work remains to elucidate whether these
plant responses are transferable to other plant species exposed to more realistic
ambient conditions.
Ozone is taken up into leaves through open stomata. Once inside the substomatal
cavity, O3 is thought to rapidly react with the aqueous layer surrounding the cell
(apoplast) to form breakdown products such as hydrogen peroxide (H2O2),
superoxide (O2 ), hydroxyl radicals (HO') and peroxy radicals (HO2'). Experimental
evidence suggests that mitogen-activated protein kinases and calcium are important
components of the signal transduction pathways, which communicate signals to the
nucleus and lead to changes in gene expression in response to O3. It is probable that
there are multiple signal transduction pathways, and their activation may depend
upon the plant species, its developmental stage and/or O3 exposure conditions.
Initiation of signal transduction pathways in O3 treated plants has also been observed
in stomatal guard cells. Reductions in stomatal conductance have been described for
many plant species exposed to O3. Some recent studies have also reported sluggish
stomatal responses and increased stomatal conductance in some situations. New
experimental evidence suggests that these effects on stomates may be due not only to
a decrease in carboxylation efficiency, but also to a direct impact of O3 on stomatal
guard cell function, leading to a changes in stomatal conductance.
Alterations in gene transcription that have been observed in O3-treated plants are
now evaluated more comprehensively using DNA microarray studies, which measure
changes in the entire transcriptome rather than measuring the transcript levels of
individual genes. These studies have demonstrated very consistent trends, even
though O3 exposure conditions (concentration, duration of exposure), plant species
and sampling times vary significantly. Genes involved in plant defense, signaling,
and those associated with the synthesis of plant hormones and secondary metabolism
are generally upregulated in plants exposed to O3, while those related to
photosynthesis and general metabolism are typically downregulated. Proteome
studies support these results by demonstrating concomitant increases or decreases in
the proteins encoded by these genes. Transcriptome analysis has also illuminated the
complex interactions that exist between several different phytohormones and how
9-34
-------
they modulate plant sensitivity and response to O3. Experimental evidence suggests
that while ethylene and salicylic acid are needed to develop O3-induced leaf lesions,
jasmonic acid acts antagonistically to ethylene and salicylic acid to limit the spread
of the lesions. Abscisic acid, in addition to its role in regulating stomatal aperture,
may also act antagonistically to the jasmonic acid signaling pathway. Changes in the
quantity and activity of these phytohormones and the interactions between them
reveal some of the complexity of plant responses to an oxidative stressor such as O3.
Another critical area of interest is to better understand and quantify the capacity of
the plant to detoxify oxygen radicals using antioxidant metabolites, such as ascorbate
and glutathione, and the enzymes that regenerate them. Ascorbate remains an
important focus of research, and, due to its location in the apoplast in addition to
other cellular compartments, it is regarded as a first line of defense against oxygen
radicals formed in the apoplast. Most studies demonstrate that antioxidant
metabolites and enzymes increase in quantity and activity in plants exposed to O3,
indicating that they play an important role in protecting plants from oxidative stress.
However, attempts to quantify the detoxification capacity of plants have remained
unsuccessful, as high quantities of antioxidant metabolites and enzymes do not
always translate into greater protection of the plant. Considerable variation exists
between plant species, different developmental stages, and the environmental and O3
exposure conditions which plants are exposed to.
As indicated earlier, the described alterations in transcript levels of genes correlate
with observed changes quantity and activity of the enzymes and metabolites involved
in primary and secondary metabolism. In addition to the generalized upregulation of
the antioxidant defense system, photosynthesis typically declines in O3 treated
plants. Declines in C fixation due to reductions in quantity and activity of Rubisco
were extensively described in the 2006 O3 AQCD. More recent studies support these
results and indicate that declines in Rubisco activity may also result from reductions
in Rubisco activase enzyme quantity. Other studies, which have focused on the light
reactions of photosynthesis, demonstrate that plant exposure to O3 results in declines
in electron transport efficiency and a decreased capacity to quench oxidizing radicals.
Therefore, the overall declines in photosynthesis observed in O3-treated plants likely
result from combined impacts on stomatal conductance, carbon fixation and the light
reactions. While photosynthesis generally declines in plants exposed to O3, catabolic
pathways such as respiration have been shown to increase. It has been hypothesized
that increased respiration may result from greater energy needs for defense and
repair. Secondary metabolism is generally upregulated in a variety of species
exposed to O3 as a part of a generalized plant defense mechanism. Some secondary
metabolites, such as flavonoids and polyamines, are of particular interest as they are
known to have antioxidant properties. The combination of decreases in
C assimilation and increases in catabolism and the production of secondary
metabolites would negatively impact plants by decreasing the energy available for
growth and reproduction.
9-35
-------
9.4 Nature of Effects on Vegetation and Ecosystems
9.4.1 Introduction
Ambient O3 concentrations have long been known to cause visible symptoms,
decreases in photosynthetic rates, decreases in growth and yield of plants as well as
many other effects on ecosystems (U.S. EPA. 2006b. 1996c. 1986. 1978a).
Numerous studies have related O3 exposure to plant responses, with most effort
focused on the yield of crops and the growth of tree seedlings. Many experiments
exposed individual plants grown in pots or soil under controlled conditions to known
concentrations of O3 for a segment of daylight hours for some portion of the plant's
life span. Information in this section also goes beyond individual plant-scale
responses to consider effects at the broader ecosystem scale, including effects related
to ecosystem services.
This section will focus mainly on studies published since the release of the 2006 O3
AQCD. However, because much O3 research was conducted prior to the 2006 O3
AQCD, the present discussion of vegetation and ecosystem response to O3 exposure
is largely based on the conclusions of the 1978, 1986, 1996, and 2006 O3 AQCDs.
9.4.1.1 Ecosystem Scale, Function, and Structure
Information presented in this section was collected at multiple spatial scales or levels
of biological organization, ranging from the physiology of a given species to
population, community, and ecosystem investigations. An ecological population is a
group of individuals of the same species and a community is an assemblage of
populations of different species interacting with one another that inhabit an area. For
this assessment, "ecosystem" is defined as the interactive system formed from all
living organisms and their abiotic (physical and chemical) environment within a
given area (TPCC, 2007a). The boundaries of what could be called an ecosystem are
somewhat arbitrary, depending on the focus of interest or study. Thus, the extent of
an ecosystem may range from very small spatial scales or levels of biological
organization to, ultimately, the entire Earth (TPCC, 2007a). All ecosystems,
regardless of size or complexity, have interactions and physical exchanges between
biota and abiotic factors, this includes both structural (e.g., soil type and food web
trophic levels) and functional (e.g., energy flow, decomposition, nitrification)
attributes.
Ecosystems can be described, in part, by their structure, i.e., the number and type of
species present. Structure may refer to a variety of measurements including the
species richness, abundance, community composition and biodiversity as well as
landscape attributes. Competition among and within species and their tolerance to
environmental stressors are key elements of survivorship. When environmental
conditions are shifted, for example, by the presence of anthropogenic air pollution,
9-36
-------
these competitive relationships may change and tolerance to stress may be exceeded.
Ecosystems may also be defined on a functional basis. "Function" refers to the suite
of processes and interactions among the ecosystem components and their
environment that involve nutrient and energy flow as well as other attributes
including water dynamics and the flux of trace gases. Plants, via such processes as
photosynthesis, respiration, C allocation, nutrient uptake and evaporation, affect
energy flow, C, nutrient cycling and water cycling. The energy accumulated and
stored by vegetation (via photosynthetic C capture) is available to other organisms.
Energy moves from one organism to another through food webs, until it is ultimately
released as heat. Nutrients and water can be recycled. Air pollution alters the function
of ecosystems when elemental cycles or the energy flow are altered. This alteration
can also be manifested in changes in the biotic composition of ecosystems.
There are at least three levels of ecosystem response to pollutants: (1) the individual
organism and its environment; (2) the population and its environment; and (3) the
biological community composed of many species and their environment (Billings,
1978). Individual organisms within a population vary in their ability to withstand the
stress of environmental change. The response of individual organisms within a
population is based on their genetic constitution, stage of growth at time of exposure
to stress, and the microhabitat in which they are growing (Levine and Pinto, 1998).
The stress range within which organisms can exist and function determines the
ability of the population to survive.
9.4.1.2 Ecosystem Services
Ecosystem structure and function may be translated into ecosystem services.
Ecosystem services are the benefits people obtain from ecosystems (UNEP, 2003).
Ecosystems provide many goods and services that are of vital importance for the
functioning of the biosphere and provide the basis for the delivery of tangible
benefits to human society. Hassan et al. (2005) define these benefits to include
supporting, provisioning, regulating, and cultural services:
• Supporting services are necessary for the production of all other ecosystem
services. Some examples include biomass production, production of
atmospheric O2, soil formation and retention, nutrient cycling, water cycling,
and provisioning of habitat. Biodiversity is a supporting service that is
increasingly recognized to sustain many of the goods and services that humans
enjoy from ecosystems. These provide a basis for three higher-level categories
of services.
• Provisioning services, such as products (Gitay et al., 2001), i.e., food
(including game, roots, seeds, nuts and other fruit, spices, fodder), water, fiber
(including wood, textiles), and medicinal and cosmetic products (such as
aromatic plants, pigments).
• Regulating services that are of paramount importance for human society such
as (1) C sequestration, (2) climate and water regulation, (3) protection from
9-37
-------
natural hazards such as floods, avalanches, or rock-fall, (4) water and air
purification, and (5) disease and pest regulation.
• Cultural services that satisfy human spiritual and aesthetic appreciation of
ecosystems and their components including recreational and other nonmaterial
benefits.
In the sections that follow, available information on individual, population and
community response to O3 will be discussed. Effects of O3 on productivity and
C sequestration, water cycling, below-ground processes, competition and
biodiversity, and insects and wildlife are considered below and in the context of
ecosystem services where appropriate.
9.4.2 Visible Foliar Injury and Biomonitoring
Visible foliar injury resulting from exposure to O3 has been well characterized and
documented over several decades on many tree, shrub, herbaceous, and crop species
(U.S. EPA. 2006b. 1996b. 1984. 1978a). Visible foliar injury symptoms are
considered diagnostic as they have been verified experimentally in exposure-
response studies, using exposure methodologies such as CSTRs, OTCs, and free-air
fumigation (see Section 9.2 for more detail on exposure methodologies). Several
pictorial atlases and guides have been published, providing details on diagnosis and
identification of O3-induced visible foliar injury on many plant species throughout
North America (Flagler. 1998: NAPAP. 1987) and Europe dimes etal.. 2001:
Sanchez et al., 2001). Typical visible injury symptoms on broad-leaved plants
include: stippling, flecking, surface bleaching, bifacial necrosis, pigmentation
(e.g., bronzing), chlorosis, and/or premature senescence. Typical visible injury
symptoms for conifers include: chlorotic banding, tip burn, flecking, chlorotic
mottling, and/or premature senescence of needles. Although common patterns of
injury develop within a species, these foliar lesions can vary considerably between
and within taxonomic groups. Furthermore, the degree and extent of visible foliar
injury development varies from year to year and site to site (Smith. 2012:
Orendovici-Best et al.. 2008: Chappelka et al.. 2007: Smith et al.. 2003). even among
co-members of a population exposed to similar O3 levels, due to the influence of co-
occurring environmental and genetic factors (Souza et al.. 2006: Chappelka et al..
2003: Somers et al.. 1998). Nevertheless, Chappelka et al. (2007) reported that the
average incidence of O3-induced foliar injury was 73% on milkweed observed in the
Great Smoky Mountains National Park in the years 1992-1996.
Although the majority of O3-induced visible foliar injury occurrence has been
observed on seedlings and small plants, many studies have reported visible injury of
mature coniferous trees, primarily in the western U.S. (Arbaugh et al.. 1998) and to
mature deciduous trees in eastern North America (Schaub et al.. 2005: Vollenweider
etal.. 2003: Chappelka et al.. 1999a: Chappelka et al.. 1999b: Somers et al.. 1998:
Hildebrand et al.. 1996).
9-38
-------
It is important to note that visible foliar injury occurs only when sensitive plants are
exposed to elevated O3 concentrations in a predisposing environment. A major
modifying factor for O3-induced visible foliar injury is the amount of soil moisture
available to a plant during the year that the visible foliar injury is being assessed.
This is because lack of soil moisture generally decreases stomatal conductance of
plants and, therefore, limits the amount of O3 entering the leaf that can cause injury
(Matvssek et al.. 2006: Panek. 2004: Grulke et al.. 2003a: Panek and Goldstein.
2001: Temple et al.. 1992: Temple et al.. 1988). Consequently, many studies have
shown that dry periods in local areas tend to decrease the incidence and severity of
Os-induced visible foliar injury; therefore, the incidence of visible foliar injury is not
always higher in years and areas with higher O3, especially with co-occurring
drought (Smith. 2012: Smith et al.. 2003). Other factors such as leaf age influence the
severity of symptom expression with older leaves showing greater injury severity as
a result of greater seasonal exposure (Zhang et al.. 2010a).
Although visible injury is a valuable indicator of the presence of phytotoxic
concentrations of O3 in ambient air, it is not always a reliable indicator of other
negative effects on vegetation. The significance of O3 injury at the leaf and whole
plant levels depends on how much of the total leaf area of the plant has been affected,
as well as the plant's age, size, developmental stage, and degree of functional
redundancy among the existing leaf area. Previous O3 AQCDs have noted the
difficulty in relating visible foliar injury symptoms to other vegetation effects such as
individual plant growth, stand growth, or ecosystem characteristics (U.S. EPA,
2006b, 1996b). As a result, it is not presently possible to determine, with consistency
across species and environments, what degree of injury at the leaf level has
significance to the vigor of the whole plant. However, in some cases, visible foliar
symptoms have been correlated with decreased vegetative growth (Somers et al..
1998: Karnoskv et al.. 1996: Peterson et al.. 1987: Benoit et al.. 1982) and with
impaired reproductive function (Chappelka. 2002: Black et al.. 2000). Conversely,
the lack of visible injury does not always indicate a lack of phytotoxic concentrations
of O3 or a lack of non-visible O3 effects (Gregg et al.. 2006. 2003).
9.4.2.1 Biomonitoring
The use of biological indicators to detect phytotoxic levels of O3 is a longstanding
and effective methodology (Chappelka and Samuelson. 1998: Manning and Krupa.
1992). A plant bioindicator can be defined as a vascular or nonvascular plant
exhibiting a typical and verifiable response when exposed to a plant stress such as an
air pollutant (Manning. 2003). To be considered a good indicator species, plants must
(1) exhibit a distinct, verified response; (2) have few or no confounding disease or
pest problems; and (3) exhibit genetic stability (U.S. EPA. 2006b). Such sensitive
plants can be used to detect the presence of a specific air pollutant such as O3 in the
ambient air at a specific location or region and, as a result of the magnitude of their
response, provide unique information regarding specific ambient air quality.
Bioindicators can be either introduced sentinels, such as the widely used tobacco
9-39
-------
(Nicotiana tabacum) variety Bel W3 (Calatayud et al., 2007b; Laffray et al., 2007;
Nali et al.. 2007: Gombert et al.. 2006: Kostka-Rick and Hahn. 2005: Heggestad
1991) or detectors, which are sensitive native plant species (Chappelka et al., 2007:
Souzaet al., 2006). The approach is especially useful in areas where O3 monitors are
not operated (Manning, 2003). For example, in remote wilderness areas where
instrument monitoring is generally not available, the use of bioindicator surveys in
conjunction with the use of passive samplers (Krupa et al.. 2001) may be a useful
methodology (Manning. 2003). However, it requires expertise in recognizing those
signs and symptoms uniquely attributable to exposure to O3 as well as in their
quantitative assessment.
Since the 2006 O3 AQCD, new sensitive plant species have been identified from
field surveys and verified in controlled exposure studies (Kline et al., 2009: Kline et
al., 2008). Several multiple-year field surveys have also been conducted at National
Wildlife Refuges in Maine, Michigan, New Jersey, and South Carolina (Davis, 2009,
2007a. b; Davis and Orendovici. 2006).
The USDA Forest Service through the Forest Health Monitoring Program (FHM)
(1990 - 2001) and currently the Forest Inventory and Analysis (FIA) Program has
been collecting data regarding the incidence and severity of visible foliar injury on a
variety of O3 sensitive plant species throughout the U.S. (Smith. 2012: Coulston et
al.. 2003: Smith et al.. 2003). The plots where these data are taken are known as
biosites. These biosites are located throughout the country and analysis of visible
foliar injury within these sites follows a set of established protocols. For more
details, see http://www.nrs.fs.fed.us/fia/topics/ozone/ (USDA. 2011). The network
has provided evidence of O3 concentrations high enough to induce visible symptoms
on sensitive vegetation. From repeated observations and measurements made over a
number of years, specific patterns of areas experiencing visible O3 injury symptoms
can be identified. (Coulston et al., 2003) used information gathered over a 6-year
period (1994-1999) from the network to identify several species that were sensitive
to O3 over entire regions, including sweetgum (Liquidambar styraciflua), loblolly
pine (Pinus taeda), and black cherry (P. serotina). A recent paper by Smith et al.
(2012) reported trends in foliar O3 injury in the northeast and north central U.S.
within the biomonitoring network over a 16-year period (1994-2009). The results
showed that incidence and severity of foliar injury were dependent upon local site
conditions (i.e., soil moisture availability) and O3 exposure. Overall, there was a
declining trend in the incidence of foliar injury as peak O3 concentrations declined.
Nevertheless, moderate O3 exposures continued to cause foliar injury at sites
throughout the region.
In a study of the west coast of the US, Campbell et al. (2007) reported O3 injury in
25-37% of biosites in California forested ecosystems from 2000-2005.
A study by Kohut et al. (2007) assessed the estimated risk of O3-induced visible
foliar injury on bioindicator plants (NFS, 2006) in 244 national parks in support of
the National Park Service's Vital Signs Monitoring Network (NFS, 2007). The risk
assessment was based on a simple model relating response to the interaction of
species, level of O3 exposure, and exposure environment. Kohut et al. (2007)
9-40
-------
concluded that the estimated risk of visible foliar injury was high in 65 parks (27%),
moderate in 46 parks (19%), and low in 131 parks (54%). Some of the well-known
parks with a high risk of O3-induced visible foliar injury include Gettysburg, Valley
Forge, Delaware Water Gap, Cape Cod, Fire Island, Antietam, Harpers Ferry,
Manassas, Wolf Trap Farm Park, Mammoth Cave, Shiloh, Sleeping Bear Dunes,
Great Smoky Mountains, Joshua Tree, Sequoia and Kings Canyon, and Yosemite.
Lichens have also long been used as biomonitors of air pollution effects on forest
health (Nash, 2008). It has been suspected, based on field surveys in the San
Bernardino Mountains surrounding the Los Angeles air basin, that declines in lichen
diversity and abundance were correlated with measured O3 gradients (Giil et al.,
2011). Several recent studies in North America (Geiser and Neitlich, 2007; Gombert
et al., 2006; Jovan and McCune, 2006) and Europe (Nali et al., 2007; Gombert et al.,
2006) have used lichens as biomonitors of atmospheric deposition (e.g., N and S) and
O3 exposure. Nali et al. (2007) found that epiphytic lichen biodiversity was not
related to O3 geographical distribution. In addition, a recent study by Riddell et al.
(2010) found that lichen species, Ramalina menziesii, showed no decline in
physiological response to low and moderate concentrations of O3 and may not be a
good indicator for O3 pollution. Mosses have also been used as biomonitors of air
pollution; however, there remains a knowledge gap in the understanding of the
effects of O3 on mosses as there has been very little information available on this
topic in recent years.
9.4.2.2 Summary
Visible foliar injury resulting from exposure to O3 has been well characterized and
documented over several decades of research on many tree, shrub, herbaceous, and
crop species (U.S. EPA. 2006K 1996K 1984. 1978a). Ozone-induced visible foliar
injury symptoms on certain bioindicator plant species are considered diagnostic as
they have been verified experimentally in exposure-response studies, using exposure
methodologies such as continuous stirred tank reactors (CSTRs), OTCs, and free-air
fumigation. Experimental evidence has clearly established a consistent association of
visible injury with O3 exposure, with greater exposure often resulting in greater and
more prevalent injury. Since the 2006 O3 AQCD, results of several multi-year field
surveys of O3-induced visible foliar injury at National Wildlife Refuges in Maine,
Michigan, New Jersey, and South Carolina have been published. New sensitive
species showing visible foliar injury continue to be identified from field surveys and
verified in controlled exposure studies.
The use of biological indicators in field surveys to detect phytotoxic levels of O3 is a
longstanding and effective methodology. The USDA Forest Service through the
Forest Health Monitoring (FHM) Program (1990-2001) and currently the Forest
Inventory and Analysis (FIA) Program has been collecting data regarding the
incidence and severity of visible foliar injury on a variety of O3 sensitive plant
species throughout the United States. The network has provided evidence that O3
9-41
-------
concentrations were high enough to induce visible symptoms on sensitive vegetation.
From repeated observations and measurements made over a number of years, specific
patterns of areas experiencing visible O3 injury symptoms can be identified. As noted
in the preceding section, a study of 244 national parks indicated that the estimated
risk of visible foliar injury was high in 65 parks (27%), moderate in 46 parks (19%),
and low in 131 parks (54%).
Evidence is sufficient to conclude that there is a causal relationship between
ambient O3 exposure and the occurrence of O3-induced visible foliar injury on
sensitive vegetation across the U.S.
9.4.3 Growth, Productivity and Carbon Storage in Natural Ecosystems
Ambient O3 concentrations have long been known to cause decreases in
photosynthetic rates, decreases in growth, and decreases in yield (U.S. EPA, 2006b,
1996c, 1986, 1978a). The O3-induced damages at the plant scale may translate to
damages at the stand, then ecosystem scales, and cause changes in productivity and C
storage. This section focuses on the responses of C cycling to seasonal or multi-year
exposures to O3 at levels of organization ranging from individual plants to
ecosystems. Quantitative responses include changes in plant growth, plant biomass
allocation, ecosystem production and ecosystem C sequestration. Most information
available on plant-scale responses was obtained from studies that used a single
species, especially tree seedlings and crops, while some used mixtures of herbaceous
species. Ecosystem changes are difficult to evaluate in natural settings, due to the
complexity of interactions, the number of potential confounders, and the large spatial
and temporal scales. The discussion of ecosystem effects focuses on new studies at
the large-scale FACE experiments and on ecological model simulations.
9.4.3.1 Plant Growth and Biomass Allocation
The previous O3 AQCDs concluded that there is strong evidence that exposure to O3
decreases photosynthesis and growth in numerous plant species (U.S. EPA, 2006b,
1996b, 1984, 1978a). Studies published since the last review support those
conclusions and are summarized below.
In general, research conducted over several decades has indicated that exposure to O3
alters stomatal conductance and reduces photosynthesis in a wide variety of plant
species. In a review of more than 55 studies, Wittig et al. (2007) reported that current
O3 concentrations in the northern hemisphere are decreasing stomatal conductance
(13%) and photosynthesis (11%) across tree species. It was also found that younger
trees (<4 years) were affected less by O3 than older trees. Further, the authors also
found that decreases in photosynthesis are consistent with the cumulative uptake of
O3 into the leaf. In contrast, several studies reported that O3 exposure may result in
loss of stomatal control, incomplete stomatal closure at night and a decoupling of
9-42
-------
photosynthesis and stomatal conductance, which may have implications for whole-
plant water use (Section 9.4.5).
In a recently published meta-analysis, Wittig et al. (2009) quantitatively compiled
peer reviewed studies from the past 40 years on the effect of current and future O3
exposures on the physiology and growth of forest species. They found that current
ambient O3 concentrations as reported in those studies significantly decreased annual
total biomass growth (7%) across 263 studies. The authors calculated the ambient O3
concentrations across these studies to average 40 ppb. This average was calculated
across the duration of each study and there were therefore many hourly exposures
well above 40 ppb. The decreased growth effect was reported to be greater (11 to
17%) in elevated O3 exposures (97 ppb) (Wittig et al., 2009). This meta-analysis
demonstrates the coherence of O3 effects across numerous studies and species that
used a variety of experimental techniques, and these results support the conclusion of
the previous AQCD that exposure to O3 decreases plant growth.
In two companion papers, McLauglin et al. (2007a; 2007b) investigated the effects of
ambient O3 on tree growth and hydrology at forest sites in the southern Appalachian
Mountains. The authors reported that the cumulative effects of ambient levels of O3
decreased seasonal stem growth by 30-50% for most tree species in a high O3 year in
comparison to a low O3 year (McLaughlin et al.. 2007a). The authors also reported
that high ambient O3 concentrations can increase whole-tree water use and in turn
reduce late-season streamflow (McLaughlin et al.. 2007b): see Section 9.4.5 for more
on water cycling.
Since the 2006 O3 AQCD, several recent studies have reported results from the
Aspen FACE "free air" O3 and CO2 exposure experiment in Wisconsin (Darbah et
al.. 2008: Riikonen et al.. 2008: Darbah et al.. 2007: Kubiske et al.. 2007: Kubiske et
al.. 2006: King et al.. 2005). At the Aspen FACE site, single-species and two-species
stands of trees were grown in 12, 30-m diameter rings corresponding to three
replications of a full factorial arrangement of two levels each of CO2 and O3
exposure. Over the first seven years of stand development, Kubiske et al. (2006)
observed that elevated O3 decreased tree heights, diameters, and main stem volumes
in the aspen community by 11, 16, and 20%, respectively. In addition, Kubiske et al.
(2007) reported that elevated O3 may change intra- and inter-species competition.
For example, O3 treatments increased the rate of conversion from a mixed aspen-
birch community to a birch dominated community. In a comparison presented in
Section 9.6.3 of this document, EPA found that effects on biomass accumulation in
aspen during the first seven years closely agreed with the exposure-response function
based on data from earlier OTC experiments.
Several studies at the Aspen FACE site also considered other growth-related effects
of elevated O3. Darbah et al. (2008: 2007) reported that O3 treatments decreased
paper birch seed weight and seed germination and that this would likely lead to a
negative impact of regeneration for that species. Riikonen et al. (2008) found that
elevated O3 decreased the amount of starch in birch buds by 16%, and reduced aspen
bud size, which may have been related to the observed delay in spring leaf
development. The results suggest that elevated O3 concentrations have the potential
9-43
-------
to alter C metabolism of overwintering buds, which may have carry-over effects in
the subsequent growing season (Riikonen et al., 2008).
Effects on growth of understory vegetation were also investigated at Aspen FACE.
Bandeff et al. (2006) found that the effects of elevated CO2 and O3 on understory
species composition, total and individual species biomass, N content, and 15N
recovery were a result of overstory community responses to those treatments;
however, the lack of apparent direct O3 treatment effects may have been due to high
variability in the data. Total understory biomass increased with increasing light and
was greatest under the open canopy of the aspen/maple community, as well as the
more open canopy of the elevated O3 treatments (Bandeff et al.. 2006). Similarly,
data from a study by Awmack et al. (2007) suggest that elevated CO2 and O3 may
have indirect growth effects on red (Trifolium pratense) and white (Trifolium repens)
clover in the understory via overstory community effects; however, no direct effects
of elevated O3 were observed.
Overall, the studies at the Aspen FACE experiment are consistent with many of the
OTC studies that were evaluated in previous O3 AQCDs demonstrating that O3
exposure decreases growth in numerous plant species. These results strengthen the
understanding of O3 effects on forests and demonstrate the relevance of the
knowledge gained from trees grown in open-top chamber studies.
For some annual species, particularly crops, the relevant measurement for an
assessment of the risk of O3 exposure is yield or growth, e.g., production of grain or
biomass. For plants grown in mixtures such as hayfields, and natural or semi-natural
grasslands (including native nonagricultural species), affected factors other than
production of biomass may be important. Such endpoints include biodiversity or
species composition, and effects on those endpoints may be indirect, resulting, for
example, from competitive interactions among plants in mixed-species communities.
Most of the available data on non-crop herbaceous species are for grasslands, with
many of the recent studies conducted in Europe. See Section 9.4.7.2 for a review of
the recent literature on O3 effects on competition and biodiversity in grasslands.
Root growth
Although O3 does not penetrate soil, it could alter root development by decreasing
C assimilation via photosynthesis leading to less C allocation to the roots (Andersen,
2003). The response of root development to O3 exposure depends on available
photosynthate within the plant and could vary over time. Many biotic and abiotic
factors, such as community dynamics and drought stress, have been found to alter
root development under elevated O3. Generally, there is clear evidence that O3
reduces C allocation to roots; however, results of a few recent individual studies have
shown negative (Jones et al., 2010), non-significant (Andersen et al., 2010; Phillips
et al., 2009) and positive effects (Pregitzer et al., 2008; Grebenc and Kraigher, 2007)
on root biomass and root: shoot ratio.
9-44
-------
An earlier study at the Aspen FACE experiment found that elevated O3 reduced
coarse root and fine roots biomass in young stands of paper birch and trembling
aspen (King et al., 2001). However, this reduction disappeared several years later.
Ozone significantly increased fine-root production (<1.0 mm) in the aspen
community (Pregitzer et al., 2008). This increase in fine root production was due to
changes in community composition, such as better survival of the O3-tolerant aspen
genotype, birch, and maple, rather than changes in C allocation at the individual tree
level (Pregitzer et al.. 2008: Zak et al.. 2007). In an adult European beech/Norway
spruce forest in Germany, drought was found to nullify the O3-driven stimulation of
fine root growth. Ozone stimulated fine-root production of beech during the humid
year, but had no significant impact on fine root production in the dry year (Matvssek
et al.. 2010: Nikolova et al.. 2010).
Using a non-destructive method, Vollsnes et al. (2010) studied the in vivo root
development of subterranean clover (Trifolium subterraneum) before, during and
after short-term O3 exposure. It was found that O3 reduced root tip formation, root
elongation, the total root length, and the ratios between below- and above-ground
growth within one week after exposure. Those effects persisted for up to three weeks;
however, biomass and biomass ratios were not significantly altered at the harvest
five weeks after exposure.
Several recent meta-analyses have generally indicated that O3 reduced C allocated to
roots. In one meta-analysis, Grantz et al. (2006) estimated the effect of O3 on the
root: shoot allometric coefficient (k), the ratio between the relative growth rate of the
root and shoot. The results showed that O3 reduced the root:shoot allometric
coefficient by 5.6%, and the largest decline of the root:shoot allometric coefficient
was observed in slow-growing plants. In another meta-analysis including 263
publications, Wittig et al. (2009) found that current O3 exposure had no significant
impacts on root biomass and rootshoot ratio when compared to pre-industrial O3
exposure. However, if O3 concentrations rose to 81-101 ppb (projected O3 levels in
2100), both root biomass and root: shoot ratio were found to significantly decrease.
Gymnosperms and angiosperms differed in their responses, with gymnosperms being
less sensitive to elevated O3. In two other meta-analyses, Wang et al. (2010) found
elevated O3 reduced biomass allocation to roots by 8.3% at ambient CO2 and 6.0% at
elevated CO2, and Morgan et al. (2003) found O3 reduced root dry weight of
soybean.
9.4.3.2 Summary
The previous O3 AQCDs concluded that there is strong and consistent evidence that
ambient concentrations of O3 decrease photosynthesis and growth in numerous plant
species across the United States . Studies published since the last review continue to
support that conclusion.
The meta-analyses by Wittig et al. (2009: 2007) demonstrate the coherence of O3
effects on plant photosynthesis and growth across numerous studies and species
9-45
-------
using a variety of experimental techniques. Furthermore, recent meta-analyses have
generally indicated that O3 reduced C allocation to roots (Wittig et al., 2009; Grantz
et al., 2006). Since the 2006 O3 AQCD, several studies were published based on the
Aspen FACE experiment using "free air," O3, and CO2 exposures in a planted forest
in Wisconsin. Overall, the studies at the Aspen FACE experiment were consistent
with many of the open-top chamber (OTC) studies that were the foundation of
previous O3 NAAQS reviews. These results strengthen the understanding of O3
effects on forests and demonstrate the relevance of the knowledge gained from trees
grown in open-top chamber studies.
Evidence is sufficient to conclude that there is a causal relationship between
ambient O3 exposure and reduced growth of native woody and herbaceous
vegetation.
9.4.3.3 Reproduction
Studies during recent decades have demonstrated O3 effects on various stages of
plant reproduction. The impacts of O3 on reproductive development, as reviewed by
Black et al. (2000), can occur by influencing (1) age at which flowering occurs,
particularly in long-lived trees that often have long juvenile periods of early growth
without flower and seed production; (2) flower bud initiation and development; (3)
pollen germination and pollen tube growth; (4) seed, fruit, or cone yields; and (5)
seed quality (Table 9-1) (U.S. EPA. 2006b). Several recent studies since the 2006 O3
AQCD further demonstrate the effects of O3 on reproductive processes in herbaceous
and woody plant species. Although there have been documented effects of O3 on
reproductive processes, a knowledge gap still exists pertaining to the exact
mechanism of these responses.
Ramo et al. (2007) exposed several meadow species to elevated O3 (40-50 ppb) and
CO2 (+100 ppm), both individually and combined, over three growing seasons in
ground-planted mesocosms, using OTCs. Elevated O3 delayed flowering of
Campanula rotundifolia and Vicia cracca. Ozone also reduced the overall number of
produced flowers and decreased fresh weight of individual Fragaria vesca berries.
Black et al. (2007) exposed Brassica campestris to 70 ppb for two days during late
vegetative growth or ten days during most of the vegetative phase. The two-day
exposure had no effect on growth or reproductive characteristics, while the 10 day
exposure reduced vegetative growth and reproductive site number on the terminal
raceme, emphasizing the importance of exposure duration and timing. Mature seed
number and weight per pod were unaffected due to reduced seed abortion, suggesting
that, although O3 affected reproductive processes, indeterminate species such as B.
campestris possess enough compensatory flexibility to avoid reduced seed
production (Black et al.. 2007).
In the determinate species, Plantago major, Black et al. (2010) found that O3 may
have direct effects on reproductive development in populations of differing
9-46
-------
sensitivity. Only the first flowering spike was exposed to 120 ppb O3 for 7 hours per
day on 9 successive days (corresponding to flower development) while the leaves
and second spike were exposed to charcoal-filtered air. Exposure of the first spike to
O3 affected seed number per capsule on both spikes even though spike two was not
exposed. The combined seed weight of spikes one and two was increased by 19% in
the two resistant populations, suggesting an overcompensation for injury, whereas, a
decrease of 21% was observed in the most sensitive population (Black et al.. 2010).
The question remains as to whether these effects are true direct O3-induced effects or
compensatory responses.
Studies by Darbah et al. (2008; 2007) of paper birch (Betula papyri/era) trees at the
Aspen FACE site in Rhinelander, WI investigated the effects of elevated O3 and/or
CO2 on reproductive fitness. Elevated O3 increased flowering, but decreased seed
weight and germination success rate of seeds from the exposed trees. These results
suggest that O3 can dramatically affect flowering, seed production, and seed quality
of paper birch, ultimately affecting its reproductive fitness (Darbah et al., 2008;
Darbah et al.. 2007).
Table 9-1 Ozone effects on plant reproductive processes.
Species
Apocynum androsaemifolium
(spreading dogbane)
Buddleia davidii
(butterfly bush)
Rubus cuneifolius
(sand blackberry)
Plantago major
(plantain)
Fragaria x ananassa
(cultivated strawberry)
Plantago major
(plantain)
Understory herbs
Condition Measures
Flowering time
Flowering time
Pollen germination
Pollen tube elongation
Fruit yield
Seed yield
Seed yield
References
Bergweileret al. (1999)
Findley et al. (1 997)
Chappelka et al. (2002)
Stewart et al. (1998)
Drogoudi and Ashmore (2001 : 2000)
Lyons and Barnes (1998): Pearson et al. (1996): Reiling and
Davison (1992): Whitfield et al. (1997)
Harward and Treshow (1 975)
Source: Derived from Table AX9-22 of the 2006 O3 AQCD.
9.4.3.4 Ecosystem Productivity and Carbon Sequestration
During the previous NAAQS review, there were limited studies that investigated the
effect of O3 exposure on ecosystem productivity and C sequestration. Recent studies
from long-term FACE experiments provide more evidence of the association of O3
exposure and changes in productivity at the ecosystem level of organization.
In addition to experimental studies, model studies also assessed the impact of O3
exposure on productivity and C sequestration from stand to global scales.
9-47
-------
In this section productivity of ecosystems is expressed in different ways depending
on the model or the measurements of a study. The most common metric of
productivity is Gross Primary Productivity. Gross Primary Productivity (GPP) is total
carbon that enters the ecosystem through photosynthesis by plants. Plants return a
larger portion of this carbon back to the atmosphere through respiration from roots
and aboveground portions of plants (Rpiant). Net primary production (NPP) is the
difference between total carbon gain (GPP) and carbon loss through Rpiant. Net
ecosystem production (NEP) is the difference between NPP and carbon loss through
heterotrophic respiration (Rhet) (mostly decomposition of dead organic matter)
(Lambers et al.. 1998). Similarly net ecosystem exchange (NEE) is the net flux of
carbon between the land and the atmosphere, typically measured using eddy
covariance techniques. Positive values of NEE usually refer to carbon released to the
atmosphere (i.e., a source), and negative values refer to carbon uptake (i.e., a sink).
Other studies have calculated net carbon exchange (NCE). NCE is defined as NPP
minus Rhet, Ec (the carbon emission during the conversion of natural ecosystems to
agriculture) and Ep (the sum of carbon emission from the decomposition of
agricultural products). For natural vegetation, Ec and Ep are equal to 0, so NCE is
equal NEP (Felzer et al.. 2005). In general, modeling studies take into account the
effect of O3 on C fixation of a system and there is generally not an effect on Rpiant,
Rhet, Ec or Ep. Therefore, decreases in GPP, NPP, NEP, NEE and NCE indicate a
general decrease in productivity of an ecosystem.
Two types of models are most often used to study the ecological consequences of O3
exposure: (1) single plant growth models such as TREGRO (Tree Growth Model)
and PnET-II (Photosynthetic EvapoTranspiration-II model) (Hogsett et al., 2008;
Martin et al., 2001; Ollinger et al., 1997b), and (2) process-based ecosystem models
such as PnET-CN, Dynamic Land Ecosystem Model (DEEM), Terrestrial Ecosystem
Model (TEM), or Met Office Surface Exchange Scheme - Top-down Representation
of Interactive Foliage and Flora Including Dynamics (MOSES-TRIFFID) (Felzer et
al.. 2009; Ren et al.. 2007b; Sitch et al.. 2007; Ollinger et al.. 2002) (Table 9-2).
In these models, carbon uptake is simulated through photosynthesis (TREGRO,
PnET -II, PnET- CN, DEEM and MOSES-TRIFFID) or gross primary production
(TEM). Photosynthesis rate at leaf level is modeled by a function of stomatal
conductance and other parameters in TREGRO, PnET -II, PnET- CN, DEEM and
MOSES-TRIFFID. Photosynthesis at canopy level is calculated by summing either
photosynthesis of different leaf types (TREGRO, DEEM, and MOSES-TRIFFID) or
photosynthesis of different canopy layers (PnET -II, PnET- CN). The detrimental
effect of O3 on plant growth is often simulated by multiplying photosynthesis rate by
a coefficient that is dependent on stomatal conductance and cumulative O3 uptake
(Table 9-2). Different plant functional groups (PFGs, such as deciduous trees,
coniferous trees or crops) show different responses to O3 exposure. PnET-II, PnET-
CN, TEM, DEEM and MOSES-TRIFFID estimate this difference by modifying net
photosynthesis with coefficients that represent the O3 induced fractional reduction of
photosynthesis for each functional group. The coefficients used in PnET-II, PnET-
CN, TEM, DEEM are derived from the functions of O3 exposure (AOT40) versus
photosynthesis reduction from Reich et al. (1987) and Tjoelker et al. (1995).
The coefficients used in MOSES-TRIFFID are derived from the O3 dose-
9-48
-------
photosynthesis response function from Pleijel et al. (2004a) and Karlsson et al.
(2004), where O3 dose is estimated by a metric named CUOt (cumulative stomatal
uptake of O3). The O3 threshold of CUOt is 1.6 nmol/m2/sec for woody PFT and 5
nmol/m2/sec for grass PFT, and is different from AOT40, which has an O3 threshold
level of 40 ppb for all PFTs. Experimental and model studies on ecosystem
productivity and C sequestration at the forest stand scale as well as regional and
global scales are reviewed in the following section.
9-49
-------
Table 9-2 Comparison of models used to simulate the ecological
consequences of O3 exposure.
Model Model feature Carbon uptake Ozone effect Reference
TREGRO Hourly or daily
step, single plant
model simulating
vegetation growth
process
Leaf: leaf photosynthesis is a
function of stomatal
conductance, mesophyll
conductance and the gradient of
CO2 from atmosphere to the
mesophyll cells
Canopy: Leaf is divided into
different ages. The canopy
photosynthesis rate is the sum of
the photosynthesis of all foliage
groups
The effect of O3 on photosynthesis is
simulated by reducing mesophyll
conductance, and increasing
respiration. The degree of O3 damage
is determined by ambient O3
exposure, and the threshold O3
concentration below which O3 does
not affect mesophyll conductance and
respiration
Hogsett et al.
(2008):
Weinstein et al.
(2005):
Tingey et al. (2004)
PnET-ll
and
PnET-
CN
PnET-ll: Monthly
time-step, single
plant model
PnET-CN:
Monthly time-step,
ecosystem model
Leaf: Maximum photosynthesis
rate is determined by a function
of foliar N concentration, and
stomatal conductance is
determined by a function of the
actual rate of the photosynthesis.
Canopy: canopy is divided into
multiple, even-mass layers and
photosynthesis is simulated by a
multilayered canopy submodel
The effect of O3 on photosynthesis is
simulated by an equation of stomatal
conductance and O3 dose (AOT40).
The model assumes that
photosynthesis and stomatal
conductance remain coupled under
O3 exposure, with a reduction in
photosynthesis for a given month
causing a proportion reduction in
stomatal conductance.
Ollinger et al.
(2002: 1997b):
Pan et al. (2009)
TEM
Monthly time-step,
ecosystem model
Ecosystem: TEM is run at a
0.5*0.5 degree resolution. Each
grid cell is classified by
vegetation type and soil texture,
and vegetation and detritus are
assumed to distribute
homogeneously within grid cells.
Carbon flows into ecosystem via
gross primary production, which
is a function of maximum rate of
assimilation, photosynthetically
active radiation, the leaf area
relative to the maximum annual
leaf area, mean monthly air
temperate, and nitrogen
availability.
The direct O3 reduction on GPP is
simulated by multiplying GPP by
f(O3)t, where f(O3)t is determined by
evapotranspiration, mean stomatal
conductance, ambient AOT40, and
O3 response coefficient empirically
derived from previous publications.
DLEM
Daily time-step
ecosystem model
Leaf: photosynthesis is a
function of 6 parameters:
photosynthetic photon flux
density, stomatal conductance,
daytime temperature, the
atmospheric CO2 concentration,
the leaf N content and the length
of daytime.
Canopy: Photosynthetic rates for
sunlit leaf and shaded leaf scale
up to the canopy level by
multiplying the estimated leaf
area index
Ecosystem: GPP is the sum of
gross C fixation of different plant
function groups
The detrimental effect of O3 is
simulated by multiplying the rate of
photosynthesis by O3eff, where O3eff
is a function of stomatal conductance,
ambient AOT40, and O3 sensitive
coefficient. Ozone's indirect effect on
stomatal conductance is also
simulated, with a reduction in
photosynthesis for a given month
causing a reduction in stomatal
conductance, and therefore canopy
conductance.
Ren et al.
(2007b; 2007a);
Zhang et al.
(2007a)
9-50
-------
Model Model feature Carbon uptake
Ozone effect
Reference
MOSES- 30 minute time-
TRIFFID step, dynamic
global vegetation
model
Leaf: photosynthesis is a
function of environmental and
leaf parameters and stomatal
conductance; Stomatal
conductance is a function of the
concentration of CO2 and H2O in
air at the leaf surface and the
current rate of photosynthesis of
the leaf
Canopy: Photosynthetic rates
scale up to the canopy level by
multiplying a function of leaf area
index and PAR extinction
coefficient
Ecosystem: GPP is the sum of
gross C fixation of different plant
function groups
The effect of O3 is simulated by
multiplying the rate of photosynthesis
by F, where F depends upon stomatal
conductance, O3 exposure, a critical
threshold for O3 damage, and O3
sensitive coefficient (functional type
dependent)
Sitch et al. (2007)
Local scale
Both experimental and modeling studies have provided new information on effects of
O3 exposure at the stand or site level, i.e., at the local scale. The above- and below-
ground biomass and net primary production (NPP) were measured at the Aspen
FACE site after 7 years of O3 exposure. Elevated O3 caused 23, 13 and 14%
reductions in total biomass relative to the control in the aspen, aspen-birch and
aspen-maple communities, respectively (King et al.. 2005). At the Kranzberg Forest
FACE experiment in Germany, O3 reduced annual volume growth by 9.5 m3/ha in a
mixed mature stand of Norway spruce and European beech (Pretzsch et al.. 2010).
At the grassland FACE experiment at Alp Flix, Switzerland, O3 reduced the seasonal
mean rates of ecosystem respiration and GPP by 8%, but had no significant impacts
on aboveground dry matter productivity or growing season net ecosystem production
(NEP) (Volk et al.. 2011). Ozone also altered C accumulation and turnover in soil, as
discussed in Section 9.4.6.
Changes in forest stand productivity under elevated O3 were assessed by several
model studies. TREGRO (Table 9-2) has been widely used to simulate the effects of
O3 on the growth of several species in different regions in the United States. Hogsett
et al. (2008) used TREGRO to evaluate the effectiveness of various forms and levels
of air quality standards for protecting tree growth in the San Bernardino Mountains
of California. They found that O3 exposures at the Crestline site resulted in a mean
20.9% biomass reduction from 1980 to 1985 and 10.3% biomass reduction from
1995 to 2000, compared to the "background" O3 concentrations (O3 concentration
in Crook County, Oregon). The level of vegetation protection projected was different
depending on the air quality scenarios under consideration. Specifically, when air
quality was simulated to just meet the California 8-h average maximum of 70 ppb
and the maximum 3 months 12-h SUM06 of 25 ppm-h, annual growth reductions
were limited to 1% or less, while air quality that just met a previous NAAQS (the
9-51
-------
2nd-highest 1-h max [125 ppb]) resulted in 6-7% annual reduction in growth,
resulting in the least protection relative to background O3 (Hogsett et al., 2008).
ZELIG is a forest succession gap model, and has been used to evaluate the dynamics
of natural stand succession. Combining TREGRO with ZELIG, Weinstein et al.
(2005) simulated the effects of different O3 levels (0.5, 1.5, 1.75, and 2 times [x]
ambient) on the growth and competitive interactions of white fir and ponderosa pine
at three sites in California: Lassen National Park, Yosemite National Park, and
Crestline. Their results suggested that O3 had little impact on white fir, but greatly
reduced the growth of ponderosa pine. If current O3 concentrations continue over the
next century, ambient O3 exposure (SUM06 of 110 ppm-h) at Crestline was
predicted to decrease individual tree C budget by 10% and decrease ponderosa pine
abundance by 16%. Effects at Lassen National Park and Yosemite National Park
sites were found to be smaller because of lower O3 exposure levels (Weinstein et al.,
2005).
To evaluate the influence of interspecies competition on O3 effects, the linked
TREGRO and ZELIG modeling system was used to predict the effects of O3 over
100 years on the basal area of species in a Liriodendron tulipifera-dommated forest
in the Great Smoky Mountains National Park (Weinstein et al.. 2001). Ambient O3
was predicted to decrease individual tree C budget by 28% and reduce the basal area
of L. tulipifem by 10%, whereas a 1.5x-ambient exposure was predicted to cause a
42% decrease in the individual tree C budget and a 30% reduction in basal area.
Individual tree C balance for Acer rubrum decreased 14% and 23% under ambient
and 1.5 x-ambient exposure, respectively. Prunus serotina was predicted to have less
than a 2% decrease in tree C balance in all scenarios, but its basal area was greatly
altered by the O3 effects on the other tree species. Basal area of A. rubrum and P.
serotina was predicted to increase for some years, but then decrease by up to 30%,
depending on the scenario. The authors cautioned that the simulation results were
heavily dependent on the assumption that only three of ten species studied could
directly respond O3 exposure and the rest of the species only indirectly responded
through competitive interactions. Very different predictions of stand dynamics may
have been simulated if more species could be parameterized to directly respond to O3
exposure.
Some results from models that include competitive interactions between tree species,
such as the linked TREGRO and ZELIG modeling system, may differ from empirical
modeling based on short-term single-species O3 exposure experiments. Single
species experiments were often performed on tree seedlings for one to three years in
open top chambers (OTCs) and indoor chambers (see Section 9.2). For example,
OTC-based experiments that were used to create O3 concentration-response
relationships (discussed in Section 9.6.2) were the basis of estimated tree seedling
biomass loss reported in studies by Hogsett et al. (1997) and in the 2007 EPA Staff
Paper (2007b). These illustrative biomass loss analyses covered one or two years of
O3 exposure based on historical monitoring that was interpolated across regions of
the United States. In contrast, competition models, such as the linked TREGRO and
ZELIG modeling system, use empirical data to parameterize the simulation of growth
9-52
-------
from a seedling into mature trees in competition with other trees (Weinstein et al.,
2005; Weinstein et al., 2001). These simulations may be run for 100 years or more
and have modeled annual exposures O3 across those years. Complicated competitive
interactions emerge across many decades of the simulation. For example, long-term
competition simulations can take into account competition for space, light, water,
tree longevity, disturbance, shade tolerance as well as the differential effects of O3 on
each species. As a result, a particular species may appear to grow poorly under O3
exposure in short-term seedling studies, but may grow relatively well under long-
term model scenarios with competition added to the analysis. It is important to note
that both of these approaches provide useful information about the long and short
term affects of O3 exposure on trees forest stands. However, it is very difficult to
validate the results of the long-term simulation of the effects of O3 on forest
composition.
The effects of O3 on stand productivity and dynamics were also studied by other tree
growth or stand models, such as ECOPHYS, INTRASTAND and LINKAGES.
ECOPHYS is a functional-structural tree growth model. The model used the linear
relationship between the maximum capacity of carboxylation and O3 dose to predict
the relative effect of O3 on leaf photosynthesis (Martin et al., 2001). Simulations
with ECOPHYS found that O3 decreased stem dry matter production, stem diameter
and leaf dry matter production, induced earlier leaf abscission, and inhibited root
growth (Martin et al., 2001). INTRASTAND is an hourly time step model for forest
stand carbon and water budgets. LINKAGES is a monthly time step model
simulating forest growth and community dynamics. Linking INTRASTAND with
LINKAGES, Hanson et al. (2005) found that a simulated increase in O3
concentration in 2100 (a mean 20-ppb increase over the current O3 concentration)
yields a 35% loss of carbon (C) in the net ecosystem exchange (NEE) with respect to
the current conditions (174 g C/m2-year).
Regional and global scales
Since the publication of the 2006 O3 AQCD, there is additional evidence suggesting
that O3 exposure alters ecosystem productivity and biogeochemical cycling at the
regional scale, i.e., at scales ranging from watershed to subcontinental divisions, and
at continental and global scales. Most of those studies were conducted by using
process-based ecosystem models (Table 9-2) and are briefly reviewed in the
following sections.
Ollinger et al. (1997a) simulated the effect of O3 on hardwood forest productivity of
64 hardwood sites in the northeastern U.S. with PnET-II (Table 9-2). Their
simulations indicated that O3 caused a 3-16% reduction in NPP from 1987 to 1992
(Table 9-3) The interactive effects of O3, N deposition, elevated CO2 and land use
history on C dynamics were estimated by PnET-CN (Table 9-2) (Ollinger et al..
2002). The results indicated that O3 offset the increase in net C exchange caused by
elevated CO2 and N deposition by 13% (25.0 g C/m2-year) under agriculture site
history, and 23% (33.6 g C/m2-year) under timber harvest site history. PnET-CN was
9-53
-------
also used to assess changes in C sequestration of U.S. Mid-Atlantic temperate forest.
Pan et al. (2009) designed a factorial modeling experiment to separate the effects of
changes in atmospheric composition, historical climatic variability and land-
disturbances on the C cycle. They found that O3 acted as a negative factor, partially
offsetting the growth stimulation caused by elevated CO2 and N deposition in U.S.
Mid-Atlantic temperate forest. Ozone decreased NPP of most forest types by 7-8%.
Among all the forest types, spruce-fir forest was most resistant to O3 damage, and
NPP decreased by only 1% (Pan et al.. 2009).
Felzer et al. (2004) developed TEM 4.3 (Table 9-2) to simulate the effects of O3 on
plant growth and estimated effects of O3 on NPP and C sequestration of deciduous
trees, conifers and crops in the conterminous United States. The results indicated that
O3 reduced NPP and C sequestration in the U.S. (Table 9-3) with the largest
decreases (over 13% in some locations) in NPP occurring in the Midwest agricultural
lands during the mid-summer. TEM was also used to evaluate the magnitude of O3
damage at the global scale (Table 9-2) (Felzer et al., 2005). Simulations for the
period 1860 to 1995 show that the largest reductions in NPP and net C exchange
occurred in the mid western U.S., eastern Europe, and eastern China (Felzer et al.,
2005). DEEM (Table 9-2) was developed to simulate the detrimental effect of O3 on
ecosystems, and has been used to examine the O3 damage on NPP and
C sequestration in Great Smoky Mountains National Park (Zhang et al., 2007a),
grassland ecosystems and terrestrial ecosystems in China (Ren et al., 2007b; Ren et
al., 2007a). Results of those simulations are listed in Table 9-3.
Instead of using AOT40 as their O3 exposure metric as PnET, TEM and DEEM did,
Sitch et al. (2007) incorporated a different O3 metric named CUOt (cumulative
stomatal uptake of O3), derived from Pleijel et al. (2004a), into the MOSES-
TPJFFID coupled model (Table 9-2). In the CUOt metric, the fractional reduction of
plant production is dependent on O3 uptake by stomata over a critical threshold for
damage with this threshold level varying by plant functional type. Consistent with
previous studies, their model simulation indicated that O3 reduced global gross
primary production (GPP), C-exchange rate and C sequestration (Table 9-3).
The largest reductions in GPP and land-C storage were projected over North
America, Europe, China and India. In the model, reduced ecosystem C uptake due to
O3 damage results in additional CO2 accumulation in the atmosphere and an indirect
radiative forcing of climate change. Their simulations indicated that the indirect
radiative forcing caused by O3 (0.62-1.09 W/m2) could have even greater impact on
global warming than the direct radiative forcing of O3 (0.89 W/m2) (Sitch et al.,
2007).
Results from the various model studies presented in Table 9-3 are difficult to
compare because of the various spatial and temporal scales used. However, all the
studies showed that O3 exposure decreased ecosystem productivity and
C sequestration. These results are consistent and coherent with experimental results
obtained from studies at the leaf, plant and ecosystem scales (Sitch et al.. 2007:
Felzer et al.. 2005). Many of the models use the same underlying function to simulate
the effect of O3 exposure to C uptake. For example the functions of O3 exposure
9-54
-------
(AOT40) versus photosynthesis reduction for PnET-II, PnET-CN, TEM, DEEM
were all from Reich et al. (1987) and Tjoelker et al. (1995). Therefore, it is not
surprising that the results are similar. While these models can be improved and more
evaluation with experimental data can be done, these models represent the state of
the science for estimating the effect of O3 exposure on productivity and
C sequestration.
9.4.3.5 Summary
During the previous NAAQS reviews, there were very few studies that investigated
the effect of O3 exposure on ecosystem productivity and C sequestration. Recent
studies from long-term FACE experiments, such as Aspen FACE, SoyFACE and the
Kranzberg Forest (Germany), provide evidence of the association of O3 exposure and
reduced productivity at the ecosystem level of organization. Studies at the leaf and
plant scales show that O3 decreased photosynthesis and plant growth, which provides
coherence and biological plausibility for the decrease in ecosystem productivity.
Results across different ecosystem models, such as TREGRO, PnET, TEM and
DEEM, are consistent with the FACE experimental evidence, which show that O3
reduced productivity of various ecosystems. Productivity is measured by various
metrics such as GPP, NPP, NEP, NCE, NEE and individual tree biomass gain. All
these metrics indicate a decrease in CO2 fixation by the systems that were studied.
Although O3 generally causes negative effects on plant growth, the magnitude of the
response varies among plant communities. For example, O3 had little impact on
white fir, but greatly reduced growth of ponderosa pine in southern California
(Weinstein et al.. 2005). Ozone decreased net primary production (NPP) of most
forest types in the Mid-Atlantic region, but had small impacts on spruce-fir forest
(Pan et al.. 2009).
In addition to plant growth, other indicators that are typically estimated by model
studies include net ecosystem CO2 exchange (NEE), C sequestration, and crop yield.
Model simulations consistently found that O3 exposure caused negative impacts on
these indicators, but the severity of these impacts was influenced by multiple
interactions of biological and environmental factors. The suppression of ecosystem
C sinks results in more CO2 accumulation in the atmosphere. Globally, the indirect
radiative forcing caused by O3 exposure through lowering the ecosystem C sink
could have an even greater impact on global warming than the direct radiative
forcing of O3 (Sitch et al.. 2007). Ozone could also affect regional C budgets through
interacting with multiple factors, such as N deposition, elevated CO2 and land use
history. Model simulations suggested that O3 partially offset the growth stimulation
caused by elevated CO2 and N deposition in both Northeast- and Mid-Atlantic-region
forest ecosystems of the U.S. (Pan et al.. 2009: Ollinger et al.. 2002).
The evidence is sufficient to infer that there is a causal relationship between O3
exposure and reduced productivity, and a likely causal relationship between O3
exposure and reduced carbon sequestration in terrestrial ecosystems.
9-55
-------
Table 9-3 Modeled effects of O3 on primary production, C exchange,
and C sequestration.
GPP
NPP
C exchange
C sequestration
Scale
Global
Global
U.S.
U.S.
Northeastern
U.S.
U.S. Mid-
Atlantic
China
Global
Global
Global
U.S.
GSM National
Park
China
China
China
Model
MOSES-
TRIFFID
TEM
TEM
TEM
PnET
PnET
DLEM
TEM
MOSES-
TRIFFID
MOSES-
TRIFFID
TEM
DLEM
DLEM
DLEM
DLEM
Index
cuota
AOT40
AOT40
AOT40
AOT40
AOT40
AOT40
AOT40
cuot
cuot
AOT40
AOT40
AOT40
AOT40
AOT40
Os Impacts
Decreased by 14-23% over the period 1901-2100
Decreased by 0.8% without agricultural
management and a decrease of 2.9% with
optimal agricultural management
Reduced by 2.3% without optimal N fertilization
and 7.2% with optimal N fertilization from
1983-1993
Reduced by 2.6-6.8% during the late
1980s to early 1990s.
A reduction of 3-16% from 1987-1992
Decreased NPP of most forest types by 7-8%
Reduced NPP of grassland in China by 8.5 Tgb C
from 1 960s to 1 990s
Reduced net C exchange (1950-1995) by 0.1 Pg
C/yr without agricultural management and 0.3 Pg
C/yrwith optimal agricultural management
Decreased global mean land-atmosphere
C fluxes by 1 .3 Pg C/yr and 1 .7 Pg C/yr for the
'high' and 'low' plant O3 sensitivity models,
respectively
Reduced land-C storage accumulation by
between 143 Pg C/yr and 263 Pg C/yr from
1900-2100
Reduced C sequestration by 1 8-38 Tg C/yr from
1950 to 1995
Decreased the ecosystem C storage of
deciduous forests by 2.5% and pine forest by
1.4% from 1971 to 2001
Reduced total C storage by 0.06% in 1960s and
1.6% in 1990s in China's terrestrial ecosystems
O3 exposure reduced the net C sink of China's
terrestrial ecosystem by 7% from 1961 to 2005
Ozone induced net carbon exchange reduction
ranged from 0.4-43.1 %, depending on different
forest type
Reference
Sitch et al. (2007)
Felzeretal. (2005)
Felzeretal. (2005)
Felzeretal. (2004)
Ollinger et al.
(1997a)
Pan et al. (2009)
Ren et al. (2007a)
Felzeretal. (2005)
Sitch et al. (2007)
Sitch et al. (2007)
Felzeretal. (2004)
Zhang et al. (2007a)
Ren et al. (2007b)
Tian et al. (201 1 )
Renetal. (2011)
aCUOt is defined as the cumulative stomatal uptake of O3, using a constant O3-uptake rate threshold oft nmol/m /sec.
bPg equals 1 xio15 grams.
9-56
-------
9.4.4 Crop Yield and Quality in Agricultural Systems
The detrimental effect of O3 on crop production has been recognized since the 1960s
and a large body of research has stemmed from that recognition. Previous O3
AQCDs have extensively reviewed this body of literature. Table 9-4 summarizes
recent experimental studies of O3 effects on agricultural crops, exclusive of growth
and yield. Growth and yield results are summarized in Table 9-17.
The actual concentration and duration threshold for O3 damage varies from species
to species and sometimes even among genotypes of the same species (Guidi et al..
2009: Sawada and Kohno. 2009: Biswas et al. 2008: Arivaphanphitak et al. 2005:
Dalstein and Vas. 2005: Keutgen et al.. 2005). A number of comprehensive reviews
and meta-analyses have recently been published discussing both the current
understanding of the quantitative effects of O3 concentration on a variety of crop
species and the potential focus areas for biotechnological improvement to a future
growing environment that will include higher O3 concentrations (Bender and Weigel.
2011: Booker et al.. 2009: Van Dingenen et al.. 2009: Ainsworth. 2008: Feng et al..
2008b: Haves et al.. 2007: Mills et al.. 2007a: Grantz et al.. 2006: Morgan et al..
2003). Since the 2006 O3 AQCD (U.S. EPA. 2006b). exposure-response indices for a
variety of crops have been suggested (Mills et al.. 2007a) and many reports have
investigated the effects of O3 concentration on seed or fruit quality to extend the
knowledge base beyond yield quantity. This section will outline the key findings
from these papers as well as highlight some of the recent research addressing the
endpoints such as yields and crop quality.
This section will also highlight recent literature that focuses on O3 damage to crops
as influenced by other environmental factors. Genetic variability is not the only
factor that determines crop response to O3 damage. Ozone concentration throughout
a growing-season is not homogeneous and other environmental conditions such as
elevated CO2 concentrations, drought, cold or nutrient availability may alleviate or
exacerbate the oxidative stress response to a given O3 concentration.
9.4.4.1 Yield
It is well known that yield is negatively impacted in many crop species in response to
high O3 concentration. However, the concentrations at which damage is observed
vary from species to species. Numerous analyses of experiments conducted in OTCs
and with naturally occurring gradients demonstrate that the effects of O3 exposure
also vary depending on the growth stage of the plant; plants grown for seed or grain
are often most sensitive to exposure during the seed or grain-filling period (Soja et
al.. 2000: Pleiiel et al.. 1998: Younglove et al.. 1994: Lee et al.. 1988a). AX9.5.4.1 of
the 2006 O3 AQCD summarized many previous studies on crop yield.
9-57
-------
Field studies and meta-analyses
The effect of O3 exposure on U.S. crops remains an important area of research and
several studies have been published on this topic since the 2006 O3 AQCD (U.S.
EPA, 2006b) (Table 9-4 and Table 9-17). For example, one study with cotton in a
crop-weed interaction study (Grantz and Shrestha. 2006) utilizing OTCs suggests
that 12-hour average O3 concentrations of 79.9 ppb decreased cotton biomass by
25% and 12-hour average O3 concentration of 122.7 ppb decreased cotton biomass
by 75% compared to charcoal filtered control (12-h avg: 12.8 ppb). Further, this
study suggests that the weed, yellow nutsedge, was less sensitive to increasing O3
concentration, which would increase weed competition (Grantz and Shrestha, 2006).
In a study of peanuts in North Carolina, near ambient and elevated exposures of O3
reduced photosynthesis and yield compared to very low O3 conditions (Booker et al.,
2007; Burkey et al., 2007). In another study, Grantz and Vu (2009) reported that
sugarcane biomass growth significantly declined under O3 exposure.
The average yield loss reported across a number of meta-analytic studies have been
published recently for soybean (Morgan et al.. 2003). wheat (Feng et al.. 2008b). rice
(Ainsworth. 2008). semi-natural vegetation (Hayes et al.. 2007). potato, bean and
barley (Feng and Kobayashi. 2009). Meta-analysis allows for the objective
development of a quantitative consensus of the effects of a treatment across a wide
body of literature. Further, this technique allows for a compilation of data across a
range of O3 fumigation techniques, durations and concentrations in order to assemble
the existing literature in a meaningful manner.
Morgan et al. (2003) reported an average seed yield loss for soybean of 24%
compared to charcoal filtered air across all O3 concentrations used in the 53
compiled studies. The decrease in seed yield appeared to be the product of nearly
equal decreases (7-12%) in seed weight, seed number and pod number. As would be
expected, the lowest O3 concentration (30-59 ppb) resulted in the smallest yield
losses, approximately 8%, while the highest O3 concentration (80-120 ppb) resulted
in the largest yield losses, approximately 35% (Morgan et al.. 2003). Further, the
oil/protein ratio within the soybean seed was altered due to growth at elevated O3
concentrations, with a decrease in oil content. The studies included in this meta-
analysis all used enclosed fumigation systems or growth chambers which may have
altered the coupling of the atmosphere to the lower plant canopy (McLeod and Long.
1999). although the results of Morgan et al. (2006). Betzelberger (2010). and the
comparisons presented in Section 9.6.3 strongly suggest that decreases in yield
between ambient and elevated exposures are not affected by exposure method.
Utilizing the Soybean Free Air gas Concentration Enrichment Facility (SoyFACE;
www.soyface.illinois.edu). Morgan et al. (2006) reported a 20% seed yield loss due
to a 23% increase in average daytime O3 concentration (56-69 ppb) within a single
soybean cultivar across two growing seasons in Illinois, which lies within the range
predicted by the meta-analysis. A further breakdown of the effects of current O3
concentrations (AOT40 of 4.7 ppm-h) on bean seed quality (Phaseolus vulgaris) has
identified that growth at current O3 concentrations compared to charcoal-filtered air
raised total lipids, total crude protein and dietary fiber content (Iriti et al.. 2009).
9-58
-------
An increase in total phenolics was also observed, however the individual phenolic
compounds responded differently, with significant decreases in anthocyanin content.
The seeds from ambient O3 exposed plants also displayed increased total antioxidant
capacity compared to charcoal-filtered air controls (Iriti et al, 2009). Betzelberger et
al. (2010) has recently utilized the SoyFACE facility to compare the impact of
elevated O3 concentrations across 10 soybean cultivars to investigate intraspecific
variability of the O3 response to find physiological or biochemical markers for
eventual O3 tolerance breeding efforts (Betzelberger et al.. 2010). They report an
average 17% decrease in yield across all 10 cultivars across two growing seasons due
to a doubling of ambient O3 concentrations, with the individual cultivar responses
ranging from -7% to -36%. The exposure-response functions derived for these 10
current cultivars were similar to the response functions derived from the NCLAN
studies conducted in the 1980s (Heagle. 1989). suggesting there has not been any
selection for increased tolerance to O3 in more recent cultivars. More complete
comparisons between yield predictions based on data from cultivars used in NCLAN
studies, and yield data for modern cultivars from SoyFACE are reported in Section
9.6.3 of this document. They confirm that the response of soybean yield to O3
exposure has not changed in current cultivars.
A meta-analysis has also been performed on studies investigating the effects of O3
concentrations on wheat (Feng et al., 2008b). Across 23 studies included, elevated
O3 concentrations (ranging from a 7-h daily average of 31-200 ppb) decreased grain
yield by 29%. Winter wheat and spring wheat did not differ in their responses;
however the response in both varieties to increasing O3 concentrations resulted in
successively larger decreases in yield, from a 20% decrease in 42 ppb to 60% in
153 ppb O3. These yield losses were mainly caused by a combination of decreases in
individual grain weight (-18%), ear number per plant (-16%), and grain number per
ear (-11%). Further, the grain starch concentration decreased by 8% and the grain
protein yield decreased by 18% due to growth at elevated O3 concentrations as well.
However, increases in grain calcium and potassium levels were reported (Feng et al..
2008b).
A recent meta-analysis found that growth at elevated O3 concentrations negatively
impacts nearly every aspect of rice performance as well (Ainsworth, 2008). While
rice is not a major crop in the U.S., it provides a staple food for over half of the
global population (IRRI, 2002) and the effects of rising O3 concentrations on rice
yields merit consideration. On average, rice yields decreased 14% in 62 ppb O3
compared to charcoal-filtered air. This yield loss was largely driven by a 20%
decrease in grain number (Ainsworth, 2008).
Feng and Kobayashi (2009) have recently compiled yield data for six major crop
species, potato, barley, wheat, rice, bean and soybean and grouped the O3 treatments
used in those studies into three categories: baseline O3 concentrations (<26 ppb),
current ambient 7- or 12-h daily O3 concentrations (31-50 ppb), and future ambient
7- or 12-h daily O3 concentrations (51-75 ppb). Using these categories, they have
effectively characterized the effects of current O3 concentrations and the effects of
future O3 concentrations compared to baseline O3 concentrations. At current O3
9-59
-------
concentrations, which ranged from 41-49 ppb in the studies included, soybean
(-7.7%), bean (-19.0%), barley (-8.9%), wheat (-9.7%), rice (-17.5%) and potato
(-5.3%) all had yield losses compared to the baseline O3 concentrations (<26 ppb).
At future O3 concentrations, averaging 63 ppb, soybean (-21.6%), bean (-41.4%),
barley (-14%), wheat (-28%), rice (-17.5%) and potato (-11.9%) all had significantly
larger yield losses compared to the losses at current O3 concentrations (<26 ppb)
(Feng and Kobavashi. 2009).
A review of OTC studies has determined the AOT40 critical level that causes a 5%
yield reduction across a variety of agricultural and horticultural species (Mills et al.,
2007a). The authors classify the species studied into three groups: sensitive,
moderate and tolerant. The sensitive crops, including watermelon, beans, cotton,
wheat, turnip, onion, soybean, lettuce, and tomato, respond with a 5% reduction in
yield under a 3-month AOT40 of 6 ppm-h. Watermelon was the most sensitive with a
critical level of 1.6 ppm-h. The moderately sensitive crops, including sugar beet,
oilseed rape, potato, tobacco, rice, maize, grape and broccoli, responded with a 5%
reduction in yield between 8.6 and 20 ppm-h. The crops classified as tolerant,
including strawberry, plum and barley, responded with a 5% yield reduction between
62-83.3 ppm-h (Mills et al.. 2007a).
Feng and Kobayashi (2009) compared their exposure-response results to those
published by Mills et al. (2007a) and found the ranges of yield loss to be similar for
soybean, rice and bean. However, Feng and Kobayashi (2009) reported smaller yield
losses for potato and wheat and larger yield losses for barley compared to the dose-
response functions published by Mills et al. (2007a), which they attributed to their
more lenient criteria for literature inclusion.
While the studies investigating the impact of various O3 concentrations on yield are
important and aid in determining the vulnerability of various crops to a variety of O3
concentrations, there is still uncertainty as to how these crops respond under field
conditions with interacting environmental factors such as temperature, soil moisture,
CO2 concentration, and soil fertility (Booker et al.. 2009). Further, there appears to
be a distinct developmental and genotype dependent influence on plant sensitivity to
O3 that has yet to be fully investigated across O3 concentrations in a field setting.
The potentially mitigating effect of breeding selection for O3 resistance has received
very little attention in the published scientific literature. Anecdotal reports suggest
that such selection may have occurred in recent decades for some crops in areas of
the country with high ambient exposures. However, the only published literature
available is on soybean and these studies indicate that sensitivity has not changed in
cultivars of soybean between the 1980s and the 2000s (Betzelberger et al.. 2010).
This conclusion for soybeans is confirmed by comparisons presented in Section 9.6.3
of this document.
Yield loss at regional and global scales
Because O3 is heterogeneous in both time and space and O3 monitoring stations are
predominantly near urban areas, the impacts of O3 on current crop yields at large
9-60
-------
geographical scales are difficult to estimate. Fishman et al. (2010) have used satellite
observations to estimate O3 concentrations in the contiguous tri-state area of Iowa,
Illinois and Indiana and have combined that information with other measured
environmental variables to model the historical impact of O3 concentrations on
soybean yield across the 2002-2006 growing seasons. When soybean yield across
Iowa, Indiana and Illinois was modeled as a function of seasonal temperature, soil
moisture and O3 concentrations, O3 had the largest contribution to the variability in
yield for the southern-most latitudes included in the dataset. Fishman et al. (2010)
determined that O3 concentrations significantly reduced soybean yield by 0.38 to
1.63% for every additional ppb of exposure across the 5 years. This value is
consistent with previous chamber studies (Heagle. 1989) and results from SoyFACE
(Morgan et al.. 2006). Satellite estimates of tropospheric O3 concentrations exist
globally (Fishman et al.. 2008). therefore utilizing this historical modeling approach
is feasible across a wider geographical area, longer time-span and perhaps for more
crop species.
The detrimental effects of O3 on crop production at regional or global scales were
also assessed by several model studies. Two large scale field studies were conducted
in the U.S. (NCLAN) and in Europe (European Open Top Chamber Programme,
EOTCP) to assess the impact of O3 on crop production. Ozone exposure-response
regression models derived from the two programs have been widely used to estimate
crop yield loss (Avnery et al., 2011 a, b; Van Dingenen et al., 2009; Tong and
Mauzerall, 2008; Wang and Mauzerall, 2004). Those studies found that O3 generally
reduced crop yield and that different crops showed different sensitivity to O3
pollution (Table 9-5). Ozone was calculated to induce a possible 45-82 million
metric tons loss for wheat globally. Production losses for rice, maize and soybean
were on the order of 17-23 million metric tons globally (Van Dingenen et al.. 2009).
The largest yield losses occur in high-production areas exposed to high O3
concentrations, such the Midwest and the Mississippi Valley regions in the U.S.,
Europe, China and India (Van Dingenen et al.. 2009; Tong et al.. 2007).
9.4.4.2 Crop Quality
In general, it appears that increasing O3 concentrations above current ambient
concentrations can cause species-dependent biomass losses, decreases in root
biomass and nutritive quality, accelerated senescence and shifts in biodiversity.
A study conducted with highbush blackberry has demonstrated decreased nutritive
quality with increasing O3 concentration despite no change in biomass between
charcoal-filtered control, ambient O3 and 2 x ambient O3 exposures (Ditchkoff et al.,
2009). A study conducted with sedge using control (30 ppb), low (55 ppb), medium
(80 ppb) and high (105 ppb) O3 treatments has demonstrated decreased root biomass
and accelerated senescence in the medium and high O3 treatments (Jones et al.,
2010). Alfalfa showed no biomass changes across two years of double ambient O3
concentrations (AOT40 of 13.9 ppm-h) using FACE fumigation (Maggio et al.,
2009). However a modeling study has demonstrated that 84% of the variability in the
9-61
-------
relative feed value in high-yielding alfalfa was due to the variability in mean O3
concentration from 1998-2002 (Lin et al.. 2007). Further, in a managed grassland
FACE system, the reduction in total biomass harvest over five years decreased twice
as fast in the elevated treatment (AOT40 of 13-59 ppm-h) compared to ambient
(AOT40 of 1-20.7 ppm-h). Compared with the ambient control, loss in annual dry
matter yield was 23% after 5 year. Further, functional groups were differentially
affected, with legumes showing the strongest negative response (Volk et al.. 2006).
However, a later study by Stampfli and Fuhrer (2010) at the same site suggested that
Volk et al. (2006) likely overestimated the effects of O3 on yield reduction because
the overlapping effects of species dynamics caused by heterogeneous initial
conditions and a change in management were not considered by these authors.
An OTC study conducted with Trifolium subterraneum exposed to filtered (<15 ppb),
ambient, and 40 ppb above ambient O3 demonstrated decreases in biomass in the
highest O3 treatment as well as 10-20% decreased nutritive quality which was mainly
attributed to accelerated senescence (Sanz et al.. 2005). A study conducted with
Eastern gamagrass and big bluestem in OTCs suggested that big bluestem was not
sensitive to O3, but gamagrass displayed decreased nutritive quality in the
2 x ambient O3 treatment, due to higher lignin content and decreased N (Lewis et al..
2006).
9.4.4.3 Summary
The detrimental effect of O3 on crop production has been recognized since the 1960s
and a large body of research has subsequently stemmed from those initial findings.
Previous O3 AQCDs have extensively reviewed this body of literature (U.S. EPA,
2006b). Current O3 concentrations across the U.S. are high enough to cause yield
loss for a variety of agricultural crops including, but not limited to, soybean, wheat,
potato, watermelon, beans, turnip, onion, lettuce, and tomato. Continued increases in
O3 concentration may further decrease yield in these sensitive crops. Despite the
well-documented yield losses due to increasing O3 concentration, there is still a
knowledge gap pertaining to the exact mechanisms of O3-induced yield loss.
Research has linked increasing O3 concentration to decreased photosynthetic rates
and accelerated senescence, which are related to yield.
Recent research has highlighted the effects of O3 on crop quality. Increasing O3
concentration decreases nutritive quality of grasses, decreases macro- and micro-
nutrient concentrations in fruits and vegetable crops, and decreases cotton fiber
quality. It is important to note that these effects, as well as those mentioned above
can occur without the expression of visible injury on the leaves. These areas of
research require further investigation to determine mechanisms and exposure-
response relationships.
During the previous NAAQS reviews, there were very few studies that estimated O3
impacts on crop yields at large geographical scales. Recent modeling studies found
that O3 generally reduced crop yield, but the impacts varied across regions and crop
9-62
-------
species. For example, the largest O3-induced crop yield losses occurred in high-
production areas exposed to high O3 concentrations, such the Midwest and the
Mississippi Valley regions of the U.S. (Van Dingenen et al., 2009). Among crop
species, the estimated yield loss for wheat and soybean were higher than for rice and
maize (Van Dingenen et al., 2009). Using satellite air-column observations with
direct air-sampling O3 data, Fishman et al. (2010) modeled the yield-loss due to O3
over the continuous tri-state area of Illinois, Iowa and Wisconsin. They determined
that O3 concentrations significantly reduced soybean yield, which further reinforces
previous results from FACE-type experiments and OTC experiments. Evidence is
sufficient to conclude that there is a causal relationship between O3 exposure
and reduced yield and quality of agricultural crops.
9-63
-------
Table 9-4
Species
Facility
Location
Alfalfa
(Medicago sativa
cv. Beaver)
Growth
chambers
Bean
(Phaseolus
vulgaris \.
cv Borlotto)
OTC, ground-
planted
Curno, Italy
Big Blue Stem
(Andropogon
gerardii)
OTC
Alabama, U.S.
Brassica napus
Growth
chambers
Belgium
Brassica napus
cv. We star
Growth
chambers
Finland
Eastern
Gamagrass
(Tripsacum
dactyloides)
OTC
Alabama, U.S.
Lettuce
(Lactuca sativa)
OTC
Carcaixent
Experimental
Station, Spain
Peanut
(Arachis
hypogaea)
OTC
Raleigh, NC;
U.S.
Summary of recent studies of O$ effects on
growth and yield).
Exposure Ozone Exposure3
Duration (Additional treatment)
1 , 2 or 3 or 5 hours/day 85 ppb
4 days (Exposure duration)
4 months Seasonal AOT40:
CF = 0.5 ppm-h;
Ambient = 4.6 ppm-h
(N/A)
4 months 12-havg:
CF = 14 ppb;
Ambient = 29 ppb;
Elevated = 71 ppb
(N/A)
4 days CF&176ppb
for 4 hours/day
(N/A)
17-26 days 8-h avg:
CF& 100 ppb
(Bt/non-Bt; herbivory)
4 months 12-h avg:
CF = 14 ppb;
Ambient = 29 ppb;
Elevated = 71 ppb
(N/A)
30 days 12-h mean:
CF = 10.2 ppb;
NF = 30.1 ppb;
NF+O3 = 62.7 ppb
(4 cultivars)
Syr 12-havg:
CF = 22 ppb;
Ambient = 46 ppb;
Elevated = 75 ppb
(CO2: 375 ppm; 548 ppm;
730 ppm)
Variable(s) measured
Relative feed value
Seed lipid,
Protein content
Fiber content
Relative feed value
Glucosinolates
VOC emissions
Relative feed value
Lipid peroxidation;
Root length
Harvest biomass
crops (exclusive of
Percent (%)
change from
CFb
(% change
from ambient)
n.s.
"high variability
among treatment
groups (N/A)
+28.5 (N/A)
+7.88 (N/A)
+ 14.54 (N/A)
n.s. (n.s.)
-41 (N/A)
-30.7 (N/A);
-34 (N/A)
-17 (-12)
+77 (+38)
-22 (-14)
-40 (-10)
Reference
Muntifering et
al. (2006b)
Iritietal.
(2009)
Lewis et al.
(2006)
Gielen et al.
(2006)
Himanen et
al. (2QQ9b)
Lewis et al.
(2006)
Calatayud et
al. (2002)
Booker et al.
(2007)
9-64
-------
Species
Facility
Location
Poa pratensis
OTC
Braunschweig,
Germany
Potato
(Solanum
tuberosum
cv. Bintje)
OTC
Sweden &
Finland
Potato
(Solanum
tuberosum
cv. Indira)
Climate
chambers
Germany
Soybean
OTC
Italy
Soybean
(Glycine max
cv. 93B15)
SoyFACE
Urbana, IL; U.S.
Soybean
(Glycine max
cv. 93B15)
SoyFACE
Urbana, IL; U.S.
Soybean
(Glycine max
cv. Essex)
OTC, ground-
planted
Raleigh, NC;
U.S.
Soybean
(Glycine max
cv. Essex)
OTCs, 21 L pots
Raleigh, NC;
U.S.
Exposure Ozone Exposure3
Duration (Additional treatment)
3 yr; 8-h avg:
4-5 weeks CF+25 = 21 .7 ppb;
in the spring NF+50 = 73.1 ppb
(Competition)
2yr CF = 10ppb;
Ambient = 25 ppb);
Ambient(+) = (36 ppb);
Ambient(++) = (47 ppb)
(N/A)
8 weeks CF = 10ppb;
Ambient = 50 ppb;
2x Ambient = 100 ppb
(CO2: 400 ppm &
700 ppm)
3 yr AOT40:
CF = 0 ppm-h;
Ambient = 3.4 ppm-h;
Elevated = 9.0 ppm-h
(Well-watered &
water-stressed)
3 yr AOT40:
May-Oct Ambient = 5-22 ppm-h;
Elevated = 20-43 ppm-h
(CO2: 550 ppm;
environmental variability)
4 months 8-h avg:
Ambient = 38.5 ppb;
Elevated = 52 ppb
(Herbivory)
2yr 12-havg:
CF = 21 ppb;
1 .5x Ambient = 74 ppb
(CO2: 370 ppm &
714 ppm)
3 months 12-h avg:
CF = 18ppb);
Elevated = 72 ppb)
(CO2:367&718)
Variable(s) measured
Relative feed value
[K], [Ca], [Mg], [P], [N] per
dry weight of tubers 'dose-
response regression, report
significant positive or
negative slope with
increasing [O3]
Pathogen infestation using
percent necrosis
Daily
evapotranspiration
Photosynthesis in new
leaves,
Herbivory
defense-related
genes
Post-harvest residue
Water-use efficiency
Percent (%)
change from
CFb
(% change
from ambient) Reference
N/A(n.s.;-8) Bender etal.
(2006)
[N] [P] [Ca] n.s.; Piikki et al.
[K] & [Mg] sig + (2007)
(N/A)
+52 (n.s.) Plessl et al.
(2007)
-28 (-14) Bou Jaoude
et al. (2008a)
N/A (n.s.) Bernacchi et
al. (2006)
N/A (N/A) Casteel et al.
(2008)
N/A (-15.46) Booker etal.
(2005)
n.s. (N/A) Booker etal.
(2004b)
9-65
-------
Species
Facility
Location
Soybean
(Glycine max)
10 cultivars)
SoyFACE
Urbana, IL; U.S.
Spring Wheat
(Triticum
aestivum
cv. Minaret;
Satu;
Drabant;
Dragon)
OTCs
Belgium,
Finland,
& Sweden
Strawberry
(Fragaria x
ananassa Duch.
Cv. Korona
& Elsanta)
Growth
chambers
Bonn, Germany
Sweet Potato
Growth
Chambers
Bonn, Germany
Tomato
(Lycopersicon
esculentum)
OTC
Valencia, Spain
Trifolium repens
& Trifolium
pretense
Aspen FACE
Rhinelander, Wl;
U.S.
Exposure Ozone Exposure3
Duration (Additional treatment)
2 yr 8-h avg (ppb):
Ambient = 46.3 & 37.9;
Elevated = 82.5 & 61 .3
(Cultivar comparisons)
7 yr Seasonal AOT40s ranged
from:
0 to16 ppm-h
(N/A)
2 months 8-h avg:
CF = Oppb;
Elevated = 78 ppb
(N/A)
4 weeks 8-h avg:
CF = 0 ppb;
Ambient <40 ppb;
Elevated = 255 ppb
(N/A)
133 days 8- mean:
CF = 16.3 ppb;
NF = 30.1 ppb;
NF(+) = 83.2 ppb
(Various cultivars;
early & late harvest)
3 months 3-mo daylight avg:
Ambient = 34.8 ppb;
1 .2x Ambient = 42.23 ppb
(CO2; 560 ppm)
Variable(s) measured
Total antioxidant capacity
Seed protein content;
1 ,000-seed weight
regressed across all
experiments
Total leaf area
Tuber weight
Brix degree
Lignin;
Dry-matter
digestibility
Percent (%)
change from
CFb
(% change
from ambient)
N/A (+19)
N/A (Significant
negative
correlation)
N/A (Significant
negative
correlation)
-16 (N/A)
-14 (-11. 5)
-7.2 (-3.6)
N/A (+19.3)
N/A (-4.2)
Reference
Betzelberger
et al. (2010)
Piikki et al.
(2008b)
Keutgen et
al. (2005)
Keutgen et
al. (2008)
Dalstein et al.
(2005)
Muntifering et
al. (2006a)
aOzone exposure in ppb unless otherwise noted.
bCF = Carbon-filtered air; NF = Non-filtered air.
9-66
-------
Table 9-5 Modeled effects of O3 on crop yield loss at regional and global
scales.
Scale
Global
Global
U.S.
U.S.
East
Asia
Index
M7a;
M12b;
AOT40
M12b;
AOT40
M7;
M12;
AOT40
SUM06
M7;
M12
Os Impacts
Reduced by 7.3% to 12.3% for wheat, 5.4% to 15.6% for soybean, 2.8% to
3.7% for rice, and 2.4% to 4.1% for maize in year 2000.
Os-induced global yield reductions ranged from 8.5-14% for soybean, 3.9-
15% for wheat, and 2.2-5.5% for maize in year 2000. Global crop
production losses totaled 79-1 21 million metric tons, worth $11-18 billion
annually (in U.S. Dollars; 2000).
Reduced by 4.1 % to 4.4% for wheat, 7.1 % to 1 7.7% for soybean, 2.6% to
3.2% for rice, and 2.2% to 3.6% for maize in year 2000.
Caused a loss of 53.8 million to 438 million bushels in soybean production,
which account for 1.7-14.2% of total U.S. soybean production in 2005
Reduced the yield of wheat, rice and corn by 1-9% and soybean by 23-
27% in China, Japan and South Korea in 1990
Reference
Van Dingenen et al. (2009)
Avnery et al. (2011 a)
Van Dingenen et al. (2009)
long et al. (2007)
Wang et al. (2004)
aM7 is defined as 7-hour mean O3 concentration (ppb).
bM12 is defined as 12-hour mean O3 concentration (ppb).
9.4.5 Water Cycling
Ozone can affect water use in plants and ecosystems through several mechanisms
including damage to stomatal functioning and loss of leaf area. Figure 9-7 provides a
simple illustration of potential effects of O3 exposure on water cycling. Section 9.3.2
reviewed possible mechanisms for effects of O3 exposure on stomatal functioning.
This section on water cycling discusses how this alteration of stomatal functioning
may affect water use in leaves, whole plants, a planted forest and watersheds. .
9-67
-------
O3 exposure
Altered stomatal
conductance, sluggish
stomatal response,
^canopy leaf area loss
Altered canopy
f ^> water loss
Stream flows
Soil moisture
Figure 9-7 The potential effects of O3 exposure on water cycling.
In the literature, there is not a clear consensus on the nature of leaf-level stomatal
conductance response to O3 exposure. At the leaf level, O3 exposure is known to
result in stomatal patchiness (Paoletti and Grulke, 2005; Omasa et al., 1987; Ellenson
and Amundson. 1982), i.e., the heterogeneous aperture widths of stomata on the leaf
surface, and, as a result, the collective response of groups of stomata on leaves and
canopies determines larger-scale responses to O3. When measured at steady-state
high light conditions, leaf-level stomatal conductance is often found to be reduced
when exposed to O3. For example, a meta-analysis of 55 studies found that O3
reduced stomatal conductance by 11% (Wittig et al., 2007). However, these steady-
state measurements were generally taken at saturating light conditions and steady-
state vapor pressure deficit (VPD). Saturating light and steady-state VPD conditions
are not common in the field since many parts of the plant canopy are shaded
throughout the day. When studied under varying environmental conditions, many
9-68
-------
studies have reported incomplete stomatal closure with elevated O3 exposure during
the day (Mills et al.. 2009: Grulke et al.. 2007b: Matvssek et al.. 1995: Wieser and
Havranek, 1995) or at night (Grulke et al., 2004). This may be due to sluggish
stomatal response. Sluggish stomatal response, defined as a delay in stomatal
response to changing environmental factors relative to controls (Paoletti and Grulke,
2010) has also been documented by several researchers (Grulke et al.. 2007c:
Matvssek et al.. 1995: Pearson and Mansfield. 1993: Wallin and Skarbv. 1992: Lee et
al.. 1990: Skarbv etal. 1987: Keller and Hasler. 1984: Reich and Lassoie. 1984).
Sluggish stomatal response associated with O3 exposure suggests an uncoupling of
the normally tight relationship between carbon assimilation and stomatal
conductance as measured under steady-state conditions (Gregg et al.. 2006: Paoletti
and Grulke. 2005). Several tree and ecosystem models, such as TREGRO, PnET and
DLEM, rely on this tight relationship to simulate water and carbon dynamics.
The O3-induced impairment of stomatal control may be more pronounced for plants
growing under water stress (Wilkinson and Davies. 2010: Grulke et al.. 2007a:
Paoletti and Grulke. 2005: Bonn et al.. 2004: Kellomaki and Wang. 1997: Tioelker et
al.. 1995: Reich and Lassoie. 1984). Since leaf-level stomatal regulation is usually
assessed in a steady state rather than as a dynamic response to changing
environmental conditions, steady state measurements cannot detect sluggish stomatal
response. Because of sluggish stomatal responses, water loss from plants could be
greater or reduced under dynamic environmental conditions over days and months.
In situations where stomata fail to close under low light or water stressed conditions,
water loss may be greater over time. In other situations, it is possible that slugglish
stomata may fail to completely open in response to environmental stimuli and result
in decreased water loss.
In addition to the impacts on stomatal performance, O3-induced physiological
changes, such as reduced leaf area index and accelerated leaf senescence could alter
water use efficiency. It is well established from chamber and field studies that O3
exposure is correlated with lower foliar retention (Karnosky et al.. 2003: Topa et al..
2001: Pell etal.. 1999: Grulke and Lee. 1997: Karnoskv et al.. 1996: Miller et al..
1972: Miller et al.. 1963). However, Lee et al. (2009a) did not find changes in needle
area of ponderosa pine and reported that greater canopy conductance followed by
water stress under elevated O3 may have been caused by stomatal dysfunction. At the
Aspen FACE experiment, stand-level water use, as indicated by sap flux per unit
ground area, was not significantly affected by elevated O3 despite a 22% decrease in
leaf area index and 20% decrease in basal area diddling et al.. 2008). The same study
reported a substantial increase in maximum sap flow per unit leaf area under elevated
O3, indicating higher canopy conductance compared to controls. A subsequent study
at Aspen FACE (Uddling et al.. 2009) reported that leaf-level conductance was not
reduced by elevated O3 as observed in most short-term experiments on tree seedlings
(Wittig et al.. 2007). The mean values of leaf-level conductance were always higher
in elevated O3 compared to controls, although this increase was not always
statistically significant (Uddling et al.. 2009). The authors also reported a less
sensitive stomatal closure response to increasing vapor pressure deficit in pure aspen
stands exposed to elevated O3. This indicated that there was some evidence of
impaired stomatal control. These studies at Aspen FACE also suggested that long-
9-69
-------
term cumulative effects of elevated O3 on tree and stand structure may be more
important than the primary stomatal responses for understanding the effect of O3 on
stand-level water use (Uddling et al., 2009; 2008). Elevated O3 could also affect
evapotranspiration by altering tree crown interception of precipitation. Ozone was
shown to change branch architectural parameters, and the effects were species-
dependent at the Aspen FACE experiment (Rhea et al.. 2010). The authors found that
there was a significant correlation between canopy architecture parameters and
stemflow (the flow of intercepted water down the stem of a tree) for birch but not
aspen.
It is difficult to scale up physiology measurements from leaves to ecosystems. Thus,
the current understanding of how stomatal response at the leaf scale is integrated at
the scale of whole forest canopies, and therefore how it influences tree and forest
stand water use is limited. Field studies by (McLaughlin et al., 2007a; 2007b)
provided valuable insight into the possible consequences of stomatal sluggishness for
ecosystem water cycling. McLaughlin et al. (2007a; 2007b) indicated that O3
increased water use in a mixed deciduous forest in eastern Tennessee. McLaughlin et
al. (2007a; 2007b) found that O3, with daily maximum levels ranging from 69 to
83 ppb, reduced stem growth by 30-50% in the high-O3 year 2002. The decrease in
growth rate was caused in part by amplification of diurnal cycles of water loss and
recovery. Peak hourly O3 exposure increased the rate of water loss through
transpiration as indicated by the increased stem sap flow. The authors suggested that
a potential mechanism for the increased sap flow could be altered stomatal regulation
from O3 exposure, but this was inferred through sap flow measurements and was not
directly measured. Alternatively, stomatal conductance may have increased under
higher O3 conditions (Paoletti and Grulke, 2010). The increased canopy water loss
resulted in higher water uptake by the trees as reflected in the reduced soil moisture
in the rooting zone. The change in tree water use led to further impacts on the
hydrological cycle at the landscape level. Increased water use under high O3
exposure was reported to reduce late-season modeled streamflow in three forested
watersheds in eastern Tennessee (McLaughlin et al.. 2007b).
Felzer et al. (2009) used TEM-Hydro to assess the interactions of O3, climate,
elevated CO2 and N limitation on the hydrological cycle in the eastern United States.
They found that elevated CO2 decreased evapotranspiration by 2-4% and increased
runoff by 3-7%, as compared to the effects of climate alone. When O3 damage and
N limitation were included, evapotranspiration was reduced by an additional 4-7%
and runoff was increased by an additional 6-11% (Felzer et al., 2009). Based upon
simulation with INTRAST and LINKAGES, Hanson et al. (2005) found that
increasing O3 concentration by 20 ppb above the current ambient level yields a
modest 3% reduction in water use. Those ecological models were generally built on
the assumption that O3 induces stomatal closure and have not incorporated possible
stomatal sluggishness due to O3 exposure. Because of this assumption, results of
those models normally found that O3 reduced water use.
9-70
-------
9.4.5.1 Summary
Although the evidence was from a limited number of field and modeling studies,
findings showed an association between O3 exposure and alteration of water use and
cycling in vegetation, and at the watershed level. There is not a clear consensus on
the nature of leaf-level stomatal conductance response to O3 exposure. When
measured under steady-state high light conditions, leaf-level stomatal conductance is
often found to be reduced when plants are exposed to O3. However, measurements of
stomatal conductance under dynamic light and VPD conditions indicate sluggish
responses under elevated O3 exposure, which could potentially lead to increased
water loss from vegetation in some situations. Field studies conducted by
McLaughlin et al. (2007a; 2007b) suggested that peak hourly O3 exposure increased
the rate of water loss from several tree species, and led to a reduction in the late-
season modeled stream flow in three forested watersheds in eastern Tennessee.
Sluggish stomatal responses during O3 exposure was suggested as a possible
mechanism for increased water loss during peak O3 exposure. Currently, the
O3-induced reduction in stomatal aperture is the biological assumption for most
process-based models. Because of this assumption, results of those models normally
found that O3 reduced water loss. For example, Felzer et al. (2009) found that O3
damage and N limitation together reduced evapotranspiration and increased runoff.
Although the direction of the response differed among studies, the evidence is
sufficient to conclude that there is likely to be a causal relationship between O3
exposure and the alteration of ecosystem water cycling.
9.4.6 Below-Ground Processes
Above-ground and below-ground processes are tightly interconnected. Because roots
and soil organisms are not exposed directly to O3, below-ground processes are
affected by O3 through alterations in the quality and quantity of C supply from
photosynthates and litterfall (Andersen. 2003). Ozone can decrease leaf C uptake by
reducing photosynthesis (Section 9.3). Ozone can also increase metabolic costs by
stimulating the production of chemical compounds for defense and repair processes,
and by increasing the synthesis of antioxidants to neutralize free radicals (see Section
9.3), both of which increase the allocation of carbon for above-ground processes.
Therefore, O3 could significantly reduce the amount of C available for allocation to
below-ground by decreasing C uptake while increasing C consumption of above-
ground processes (Andersen, 2003).
Since the 2006 O3 AQCD, there is additional evidence for O3 effects on below-
ground processes. Ozone has been found to alter root growth, soil food web structure,
decomposer activities, C turnover, water cycling and nutrient flow (Figure 9-8).
Ozone effects on root development and root biomass production and soil food web
structure are reviewed in Section 9.4.3.1 and Section 9.4.9.2. respectively. The focus
9-71
-------
in this section is on the response of litter input, decomposer activities, soil
respiration, soil C formation and nutrient cycling.
CO2, H2O
CO2, H2O
Allocation of C
retention
Altered stomatal function
Mv
Altered species competition
Litter production
and chemistry
CO, release
Soil foodweb
•Bacteria
•Fungi
•Micro & marco invertebrates
Organic matter
Soil physical &
chemical properties
Note: Arrows denote C flux pathways that are affected by O3. Dashed lines indicate where the impact of O3 is suspected but
unknown.
Source: Modified from Andersen et al. (2003).
Figure 9-8 Conceptual diagram showing where O3 alters C, water and nutrient
flow in a tree-soil system, including transfer between biotic and
abiotic components below ground that influence soil physical and
chemical properties.
9.4.6.1 Litter Carbon Chemistry, Litter Nutrient and Their
Ecosystem Budgets
Consistent with previous findings, recent studies show that, although the responses
are often species-dependent, O3 tends to alter litter chemistry (U.S. EPA. 2006b).
Alterations in chemical parameters, such as changes in C chemistry and nutrient
concentrations, were observed in both leaf and root litter (Table 9-6).
9-72
-------
At the Aspen FACE site, several studies investigated litter chemistry changes
(Parsons et al., 2008; Johnson and Pregitzer, 2007; Chapman et al., 2005; Liu et al.,
2005). In both aspen and birch leaf litter, elevated O3 increased the concentrations of
soluble sugars, soluble phenolics and condensed tannins (Parsons et al., 2008; Liu et
al., 2005). Compared to other treatments, aspen litter under elevated O3 had the
highest fiber concentration, with the lowest concentration associated with the birch
litter under the same conditions (Parsons et al.. 2008). Chapman et al. (2005)
measured chemical changes in fine root litter and found that elevated O3 decreased
lignin concentration. O3-induced chemistry changes were also reported from other
experimental sites. Results from an OTC study in Finland suggested that elevated O3
increased the concentration of acid-soluble lignin, but had no significant impact on
other chemicals such as total sugars, hemicelluloses, cellulose or total lignin in the
litter of silver birch (Kasurinen et al.. 2006). Results from the free air canopy O3
exposure experiment at Kranzberg Forest showed that O3 increased starch
concentrations but had no impact on cellulose and lignin in beech and spruce leaf
litter (Aneja et al.. 2007). The effect of O3 on three antioxidants (ascorbate,
glutathione and oc-tocopherol) in fine roots of beech was also assessed at Kranzberg
Forest. The results indicated that O3 had no significant effect on oc-tocopherol and
ascorbate concentrations, but decreased glutathione concentrations in fine roots
(Haberer et al., 2008). In addition to changing C chemistry, O3 also altered nutrient
concentrations in green leaves and litter (Table 9-6).
The combined effects of O3 on biomass productivity and chemistry changes may
alter C chemicals and nutrient contents at the canopy or stand level. For example,
although O3 had different impacts on their concentrations, annual fluxes of
C chemicals (soluble sugar, soluble phenolics, condensed tannins, lipid and
hemicelluloses), macro nutrients (N, P, K and S) and micro nutrients (Mg, B, Cu and
Zn) to soil were all reduced due to lower litter biomass productivity at Aspen FACE
(Liu et al.. 2007a; Liu et al.. 2005). In a 2-year growth chamber experiment in
Germany, N content of a spruce canopy in a mixed culture and Ca content of a beech
canopy in a monoculture was increased due to elevated O3, although leaf production
was not significantly altered by O3 (Rodenkirchen et al., 2009).
9-73
-------
Table 9-6 The effect of elevated Os on leaf/litter nutrient concentrations.
Study Site
Suonenjoki Research
Station, Finland
Aspen FACE
Aspen FACE
Kranzberg Forest,
Germany
Kranzberg Forest,
Germany
Species
Silver birch
Aspen and
birch
Birch
Beech and
spruce
Beech and
spruce
O3 Concentration
Ambient: 10-60 ppb
Elevated: 2* ambient
Ambient: 50-60 ppb
Elevated:
1.5x ambient
Ambient: 50-60 ppb
Elevated:
1.5* ambient
Ambient: 9-41 ppb
Elevated: 2* ambient
Ambient: 9-41 ppb
Elevated: 2* ambient
Response
Decreased the concentration
of P, Mn, Zn and B in leaf litter
Decreased the concentrations
of P, S, Ca and Zn,
but had no impact on the
concentrations
of N, K, Mg, Mn, B and Cu in leaf
litter.
Increase N concentration in birch litter
Increased N concentration in beach
leaf, but not in spruce needle
(1) Had no significant effects on
spruce needle chemistry;
Reference
Kasurinen et al.
(2006)
Liu et al. (2007a)
Parsons et al.
(2008)
Kozovitset al.
(2005)
Rodenkirchen et
al. (2009)
Salerno, Italy
Kuopio University
Research Garden,
Finland
Holm oak
Red Clover
Non-filtered OTC:
29 ppb
Filtered OTC: 17ppb
Ambient: 25.7 ppb
Elevated:
1.5* ambient
(2) increased Ca concentration in
beech leaves in monoculture, but had
no impacts on other nutrients
O3 had no significant impacts on litter Baldantoni et al.
C, N, lignin and cellulose (2011)
concentrations
Increased the total phenolic content of Saviranta et al.
leaves and had minor effects on the (2010)
concentrations of individual phenolic
compounds
9.4.6.2 Decomposer Metabolism and Litter Decomposition
The above- and below-ground physiological changes caused by O3 exposure cascade
through the ecosystem and affect soil food webs. In the 2006 O3 AQCD, there were
very few studies on the effect of O3 on the structure and function of soil food webs,
except two studies conducted by Larson et al. (2002) and Phillips et al. (2002). Since
the last O3 AQCD, new studies have provided more information on how O3 affects
the metabolism of soil microbes and soil fauna.
Chung et al. (2006) found that the activity of the cellulose-degrading enzyme
1,4-p-glucosidase was reduced by 25% under elevated O3 at Aspen FACE.
The decrease in cellulose-degrading enzymatic activity was associated with the lower
cellulose availability under elevated O3 (Chung et al., 2006). However, a later study
at the same site, which was conducted in the 1 Oth year of the experiment, found that
O3 had no impact on cellulolytic activity in soil (Edwards and Zak, 2011). In a
lysimeter study of beech trees (Fagus sylvatica) in Germany, soil enzyme activity
was found to be suppressed by O3 exposure (Esperschutz et al., 2009; Pritsch et al.,
2009). Except for xylosidase, enzyme activities involved in plant cell wall
9-74
-------
degradation (cellobiohydrolase, beta-glucosidase and glucuronidase) were decreased
in rhizosphere soil samples under elevated O3 (2 x ambient level) (Pritsch et al.,
2009). Similarly, Chen et al. (2009) found O3 exposure, with a 3-month AOT40 of
21-44 ppm-h, decreased the microbial metabolic capability in the rhizosphere and
bulk soil of wheat, although the observed reduction in bulk soil was not significant.
Ozone-induced change in soil organisms' activities could affect litter decomposition
rates. Results of recent studies indicated that O3 slightly reduced or had no impacts
on litter decomposition (Liu et al., 2009b; Parsons et al., 2008; Kasurinen et al.,
2006) (Baldantoni et al., 2011). The responses varied among species, sites and
exposure length. Parsons et al. (2008) collected litter from aspen and birch seedlings
at Aspen FACE site, and conducted a 23-month field litter incubation starting in
1999. They found that elevated O3 had different impacts on the decomposition of
aspen and birch litter. Elevated O3 was found to reduce aspen litter decomposition.
However, O3 accelerated birch litter decomposition under ambient CO2, but reduced
it under elevated CO2 (Parsons et al., 2008). Liu et al. (2009b) conducted another
litter decomposition study at Aspen FACE from 2003 to 2006, when stand leaf area
index (LAI) reached its maximum. During the 935-day field incubation, elevated O3
was shown to reduce litter mass loss in the first year, but not in the second year. They
suggested that higher initial tannin and phenolic concentrations under elevated O3
reduced microbial activity in the first year (Liu et al., 2009b). In an OTC experiment,
Kasurinen et al. (2006) collected silver birch leaf litter from three consecutive
growing seasons and conducted three separate litter-bag incubation experiments.
Litter decomposition was not affected by O3 exposure in the first two incubations,
but a slower decomposition rate was found in the third incubation. Their principle
component analysis indicated that the litter chemistry changes caused by O3
(decreased Mn, P, B and increased C:N) might be partially responsible for the
decreased mass loss of their third incubation. In another OTC experiment, Baldantoni
et al. (2011) found that O3 significantly reduced leaf litter decomposition ofQuercus
ilex L, although litter C, N, lignin and cellulose concentrations were not altered by
O3 exposure.
9.4.6.3 Soil Respiration and Carbon Formation
Ozone could reduce the availability of photosynthates for export to roots, and thus,
indirectly increase root mortality and turnover rates. Ozone has also been shown to
reduce above-ground litter productivity and alter litter chemistry, which would affect
the quality and quantity of the C supply to soil organisms (Section 9.4.6.1).
The complex interactions among those changes make it difficult to predict the
response of soil C cycling under elevated O3. The 2006 O3 AQCD concluded that O3
had no consistent impact on soil respiration (U.S. EPA, 2006b). Ozone could
increase or decrease soil respiration, depending on the approach and timing of the
measurements. Ozone may also alter soil C formation. However, very few
experiments directly measured changes in soil organic matter content under O3
fumigation (U.S. EPA, 2006b). Recent studies on soil respiration and soil C content
9-75
-------
also found mixed responses. Most importantly, recent results from long-term
fumigation experiments, such as the Aspen FACE experiment, suggest that
ecosystem response to O3 exposure can change over time. Observations made during
the late exposure years can be inconsistent with those during the early years,
highlighting the need for caution when assessing O3 effects based on short-term
studies (Table 9-7).
Soil Respiration
Ozone has shown inconsistent impacts on soil respiration. A sun-lit
controlled-environment chamber study found that O3 had no significant effects on
soil respiration, fine root biomass or any of the soil organisms in a reconstructed
ponderosa pine/soil-litter system (Tingey et al., 2006). In an adult European
beech/Norway spruce forest at Kranzberg Forest, the free air O3 fumigation (AOT40
of 10.2-117 ppm-h) increased soil respiration under both beech and spruce during a
humid year (Nikolova et al., 2010). The increased soil respiration under beech has
been accompanied by the increase in fine root biomass and ectomycorrhizal fungi
diversity and turnover (Grebenc and Kraigher, 2007). The stimulating effect on soil
respiration disappeared under spruce in a dry year, which was associated with a
decrease in fine root production in spruce under drought. This finding suggested that
drought was a more dominant stress than O3 for spruce (Nikolova et al., 2010).
Andersen et al. (2010) labeled the canopies of European beech and Norway spruce
with CO2 depleted in 13C at the same site. They did not observe any significant
changes in soil respiration for either species.
The nearly 10 year long studies at Aspen FACE indicated that the response of soil
respiration to O3 interacted with CO2 exposure and varied temporally (Table 9-7)
(Pregitzer et al., 2008; Pregitzer et al., 2006; King et al., 2001). Ozone treatment
alone generally had the lowest mean soil respiration rates, although those differences
between control and elevated O3 were usually not significant. However, soil
respiration rates were different with O3 alone and when acting in combination with
elevated CO2. In the first five years (1998-2002), soil respiration under +CO2+O3
treatment was similar to that under control and lower than that under +CO2 treatment
(Pregitzer et al., 2006; King et al., 2001). Since 2003, +CO2+O3 treatment started to
show the greatest impact on soil respiration. Compared to elevated CO2, soil
respiration rate under +CO2+O3 treatment was 15-25% higher from 2003-2004, and
5-10% higher from 2005-2007 (Pregitzer et al.. 2008; Pregitzer et al.. 2006). Soil
respiration was highly correlated with the biomass of roots with diameters of <2 mm
and <1 mm, across plant community and atmospheric treatments. The authors
suggested that the increase in soil respiration rate may be due to +CO2+O3 increased
fine root (<1.0 mm) biomass production (Pregitzer et al., 2008).
9-76
-------
Table 9-7 The temporal variation of ecosystem responses to O$ exposure at
Aspen FACE site
Endpoint
Litter
decomposition
Fine root production
Period of
Measurement
1999-2001
2003-2006
1999
2002, 2005
1 998-1 999
Response
O3 reduced aspen litter decomposition.
However, O3 accelerated birch litter decomposition
under ambient CO2, but reduced it under elevated
C02
O3 reduced litter mass loss in the first year,
but not in the second year.
O3 had no significant impact on fine root biomass
O3 increased fine root biomass
Soil respiration under +CO2+O3 treatment was lower
than that under +CO2 treatment
Reference
Parsons et al. (2008)
Liu et al. (2009b)
King et al. (2001 )
Pregitzeret al. (2008)
King et al. (2001 )
2003-2007
Soil respiration under +CO2+O3 treatment was
5-25% higher than under elevated CO2 treatment.
Pregitzer et al. (2008: 2006)
1998-2001
Soil C formation
O3 reduced the formation rates of total soil C by 51 %
and acid-insoluble soil C by 48%
Loya et al. (2003)
2004-2008
No significant effect of O3 on the new C formed
under elevated CO2
Talhelm et al. (2009)
Changes in leaf chemistry and productivity due to O3 exposure have been shown to
affect herbivore growth and abundance (see Section 9.4.9.1). Canopy insects could
affect soil carbon and nutrient cycling through frass deposition, or altering chemistry
and quantity of litter input to the forest floor. A study at the Aspen FACE found that
although elevated O3 affected the chemistry of frass and greenfall, these changes had
small impact on microbial respiration and no effect on nitrogen leaching (Hillstrom
et al., 2010a). However, respiratory carbon loss and nitrate immobilization were
nearly double in microcosms receiving herbivore inputs than those receiving no
herbivore inputs (Hillstrom et al., 2010a).
Soil Carbon Formation
Ozone-induced reductions in plant growth can result in reduced C input to soil and
therefore soil C content (Andersen. 2003). The simulations of most ecosystem
models support this prediction (Ren et al.. 2007b: Zhang et al.. 2007a: Felzer et al..
2004). However, very few studies have directly measured soil C dynamics under
elevated O3. After the first four years of fumigation (from 1998 to 2001) at the
Aspen FACE site, Loya et al. (2003) found that forest stands exposed to both
elevated O3 and CO2 accumulated 51% less total soil C, and 48% less acid-insoluble
soil C compared to stands exposed only to elevated CO2. Soil organic carbon (SOC)
was continuously monitored at the Aspen FACE site, and the later data showed that
the initial reduction in new C formation (soil C derived from plant litter since the
start of the experiment) by O3 under elevated CO2 is only a temporary effect
(Table 9-7) (Talhelm et al., 2009). The amount of new soil C in the elevated CO2 and
9-77
-------
the combined elevated CO2 and O3 treatments has converged since 2002. There was
no significant effect of O3 on the new C formed under elevated CO2 over the last
four years of the study (2004-2008). Talhelm et al. (2009) suggested the observed
reduction in the early years of the experiment might be driven by a suppression of
C allocated to fine root biomass. During the early exposure years, O3 had no
significant impact on fine root production (King et al.. 2001). However, the effect of
O3 on fine root biomass was observed later in the experiment. Ozone increased fine
root production and the highest fine root biomass was observed under the combined
elevated CO2 and O3 treatment in the late exposure years (Table 9-7) (Pregitzer et
al.. 2006). This increase in fine root production was due to changes in community
composition, such as better survival of an O3-tolerant aspen genotype, birch and
maple, rather than changes in C allocation at the individual tree level (Pregitzer et al..
2008: Zak et al.. 2007).
9.4.6.4 Nutrient Cycling
Ozone can affect nutrient cycling by changing nutrient release from litter, nutrient
uptake by plants, and soil microbial activity. Nitrogen is the limiting nutrient for
most temperate ecosystems, and several studies examined N dynamics under
elevated O3. Nutrient mineralization from decomposing organic matter is important
for sustaining ecosystem production. Holmes et al. (2006) found that elevated O3
decreased gross N mineralization at the Aspen FACE site, indicating that O3 may
reduce N availability. Other N cycling processes, such as NH4+ immobilization, gross
nitrification, microbial biomass N and soil organic N, were not affected by elevated
O3 (Holmes etal.. 2006). Similarly, Kanerva et al. (2006) found total N, NO3-,
microbial biomass N, potential nitrification and denitrification in their meadow
mesocosms were not affected by elevated O3 (40-50 ppb). Ozone was found to
decrease soil mineral N content at SoyFACE, which was likely caused by a reduction
in plant material input and increased denitrification (Pujol Pereira et al.. 2011).
Ozone also showed small impact on other micro and macro nutrients. Liu et al.
(2007a) assessed N, P, K, S, Ca, Mg, Mn, B, Zn and Cu release dynamics at Aspen
FACE, and they found that O3 had no effects on most nutrients, except to decrease N
and Ca release from litter. These studies reviewed above suggest that soil N cycling
processes are not affected or slightly reduced by O3 exposure. However, in a
lysimeter study with young beech trees, Stoelken et al. (2010) found that elevated O3
stimulated N release from litter which was largely attributed to an enhanced
mobilization of inert nitrogen fraction.
Using the Simple Nitrogen Cycle model (SINIC), Hong et al. (2006) evaluated the
impacts of O3 exposure on soil N dynamics and streamflow nitrate flux.
The detrimental effect of O3 on plant growth was found to reduce plant uptake of N
and therefore increase nitrate leaching. Their model simulation indicated that ambient
O3 exposure increased the mean annual stream flow nitrate export by 12%
(0.042 g N/m2-year) at the Hubbard Brook Experimental Watershed from 1964-1994
(Hong et al.. 2006).
9-78
-------
9.4.6.5 Dissolved Organic Carbon and Biogenic Trace Gases
Emission
The O3-induced changes in plant growth, C and N fluxes to soil and microbial
metabolism can alter other biogeochemical cycling processes, such as soil dissolved
organic carbon (DOC) turnover and trace gases emission.
Jones et al. (2009) collected fen cores from two peatlands in North Wales, UK and
exposed them to one of four levels of O3 (AOT40 of 0, 3.69, 5.87 and 13.80 ppm-h
for 41 days). They found the concentration of porewater DOC in fen cores was
significantly decreased by increased O3 exposure. A reduction of the low molecular
weight fraction of DOC was concurrent with the observed decrease in DOC
concentration. Their results suggested that O3 damage to overlying vegetation may
decrease utilizable C flux to soil. Microbes, therefore, have to use labile C in the soil
to maintain their metabolism, which, the authors hypothesized, leads to a decreased
DOC concentration with a shift of the DOC composition to more aromatic, higher
molecular weight organic compounds.
Several studies since the 2006 O3 AQCD have examined the impacts of O3 on
nitrous oxide (N2O) and methane (CH4) emission. Kanerva et al. (2007) measured
the fluxes of N2O and CH4 in meadow mesocosms, which were exposed to elevated
CO2 and O3 in OTCs in south-western Finland. They found that the daily N2O fluxes
were decreased in the NF+O3 (non-filtered air + elevated O3, 40-50 ppb) after three
seasons of exposure. Elevated O3 alone or combined with CO2 did not have any
significant effect on the daily fluxes of CH4 (Kanerva et al.. 2007). In another study
conducted in central Finland, the 4 year open air O3 fumigation (AOT40 of 20.8-
35.5 ppm-h for growing season) slightly increased potential CH4 oxidation by 15%
in the peatland microcosms, but did not affect the rate of potential CH4 production or
net CH4 emissions, which is the net result of the potential CH4 production and
oxidation (Morsky et al., 2008). However, several studies found that O3 could
significantly reduce CH4 emission. Toet et al. (2011) exposed peatland mesocosms to
O3 in OTCs for two years, and found that CH4 emissions were significantly reduced
by about 25% during midsummer periods of both years. In an OTC study of rice
paddy, Zheng et al. (2011) found that the daily mean CH4 emissions were
significantly lower under elevated O3 treatments than those in charcoal-filtered air
and nonfiltered air treatments. They found that the seasonal mean CH4 emissions
were negatively related with AOT40, but positively related to the relative rice yield,
aboveground biomass and underground biomass.
9.4.6.6 Summary
Since the 2006 O3 AQCD, more evidence has shown that although the responses are
often site specific, O3 altered the quality and quantity of litter input to soil, microbial
community composition, and C and nutrient cycling. Biogeochemical cycling of
below-ground processes is fueled by C input from plants. Studies at the leaf and plant
9-79
-------
level have provided biologically plausible mechanisms, such as reduced
photo synthetic rates, increased metabolic cost, and reduced root C allocation for the
association of O3 exposure and the alteration of below-ground processes.
Results from Aspen FACE and other experimental studies consistently found that O3
reduced litter production and altered C chemistry, such as soluble sugars, soluble
phenolics, condensed tannins, lignin, and macro/micro nutrient concentration in litter
(Parsons et al, 2008; Kasurinen et al., 2006; Liu et al, 2005). Under elevated O3, the
changes in substrate quality and quantity could alter microbial metabolism and
therefore soil C and nutrient cycling. Several studies indicated that O3 suppressed
soil enzyme activities (Pritsch et al., 2009; Chung et al., 2006). However, the impact
of O3 on litter decomposition was inconsistent and varied among species, sites and
exposure length. Similarly, O3 had inconsistent impacts on dynamics of micro and
macro nutrients.
Studies from the Aspen FACE experiment suggested that the response of below-
ground C cycle to O3 exposure, such as litter decomposition, soil respiration and soil
C content, changed over time. For example, in the early part of the experiment
(1998-2003), O3 had no impact on soil respiration but reduced the formation rates of
total soil C under elevated CO2. However, after 10-11 years of exposure, O3 was
found to increase soil respiration but have no significant impact on soil C formation
under elevated CO2.
The evidence is sufficient to infer that there is a causal relationship between O3
exposure and the alteration of below-ground biogeochemical cycles.
9.4.7 Community Composition
The effects of O3 on species competition (AX9.3.3.4) and community composition
(AX9.6.4) were summarized in the 2006 O3 AQCD. Plant species differ in their
sensitivity to O3. Further, different genotypes of a given species also vary in their
sensitivity. This differential sensitivity could change the competitive interactions that
lead to loss in O3 sensitive species or genotypes. In addition, O3 exposure has been
found to alter reproductive processes in plants (see Section 9.4.3.3). Changes in
reproductive success could lead to changes in species composition. However, since
ecosystem-level responses result from the interaction of organisms with one another
and with their physical environment, it takes longer for a change to develop to a level
of prominence at which it can be identified and measured. A shift in community
composition in forest and grassland ecosystems noted in the 2006 O3 AQCD has
continued to be observed from experimental and gradient studies. Additionally,
research since the last review has shown that O3 can alter community composition
and diversity of soil microbial communities.
9-80
-------
9.4.7.1 Forest
In the San Bernardino Mountains in southern California, O3 pollution caused a
significant decline in ponderosa pine (Pinus ponderosd) and Jeffrey pine (Pinus
jeffreyi) (U.S. EPA, 2006b). Pine trees in the young mature age class group exhibited
higher mortality rates compared with mature trees at a site with severe O3 visible
foliar injury. The vulnerability of young mature pines was most likely caused by the
fact that trees in this age class were emerging into the canopy, where higher O3
concentrations were encountered (McBride and Laven. 1999). Because of the loss of
O3-sensitive pines, mixed forests of ponderosa pine, Jeffery Pine and white fir (Abies
concolor) shifted to predominantly white fir (Miller, 1973). Ozone may have
indirectly caused the decline in understory diversity in coniferous forests in the San
Bernardino Mountains through an increase in pine litterfall. This increase in litterfall
from O3 exposure results in an understory layer that may prohibit the establishment
of native herbs, but not the exotic annual Galium aparine (Allen et al., 2007).
Ozone damage to conifer forests has also been observed in several other regions.
In the Valley of Mexico, a widespread mortality of sacred fir (Abies religiosd) was
observed in the heavily polluted area of the Desierto de los Leones National Park in
the early 1980s (de Lourdes de Bauer and Hernandez-Tejeda. 2007: Fenn et al..
2002). Ozone damage was widely believed to be an important causal factor in the
dramatic decline of sacred fir. In alpine regions of southern France and the
Carpathians Mountains, O3 was also considered as the major cause of the observed
decline in cembran pine (Pinus cembrd) (Wieser et al.. 2006). However, many
environmental factors such as light, temperature, nutrient and soil moisture, and
climate extremes such as unusual dry and wet periods could interact with O3 and
alter the response of forest to O3 exposure. For those pollution gradient studies,
several confounding factors, such as drought, insect outbreak and forest
management, may also contribute to or even be the dominant factors causing the
mortality of trees (de Lourdes de Bauer and Hernandez-Tejeda. 2007; Wieser et al..
2006).
Recent evidence from long-term free O3 fumigation experiments provided additional
support for the potential impacts of O3 on species competition and community
composition changes in forest ecosystems. At the Aspen FACE site, community
composition at both the genetic and species levels was altered after seven years of
fumigation with O3 (Kubiske et al.. 2007). In the pure aspen community, O3
fumigation reduced growth and increased mortality of sensitive clone 259, while the
O3 tolerant clone 8L emerged as the dominant clone. Growth of clone 8L was even
greater under elevated O3 compared to controls, probably due to O3 alleviated
competitive pressure on clone 8L by reducing growth of other clones. In the mixed
aspen-birch and aspen-maple communities, O3 reduced the competitive capacity of
aspen compared to birch and maple (Kubiske et al.. 2007). In a phytotron study, O3
fumigation reduced growth of beech but not spruce in mixed culture, suggesting a
higher susceptibility of beech to O3 under interspecific competition (Kozovits et al..
2005).
9-81
-------
9.4.7.2 Grassland and Agricultural Land
The response of managed pasture, often cultivated as a mixture of grasses and clover,
to O3 pollution has been studied for many years. The tendency for O3-exposure to
shift the biomass of grass-legume mixtures in favor of grass species, reported in the
previous O3 AQCD has been generally confirmed by recent studies. In a mesocosm
study, Trifolium repens and Lolium perenne mixtures were exposed to an episodic
rural O3 regime within solardomes for 12 weeks. T. repens showed significant
changes in biomass but not L. perenne, and the proportion of T. repens decreased in
O3-exposed mixtures compared to the control (Haves et al, 2009). The changes in
community composition of grass-legume-forb mixtures were also observed at the Le
Mouret FACE experiment, Switzerland. During the 5-year O3 fumigation (AOT40 of
13.3-59.5 ppm-h), the dominance of legumes in fumigated plots declined more
quickly than those in the control plots (Volk et al., 2006). However, Stampfli and
Fuhrer (2010) reanalyzed the species and soil data and suggested that Volk et al.
(2006) overestimated the O3 effect. Stampfli and Fuhrer (2010) found that the
difference in the species dynamics between control and O3 treatment was more
caused by heterogeneous initial conditions than O3 exposure. Several studies also
suggested that mature/species-rich ecosystems were more resilient to O3 exposure.
At another FACE experiment, located at Alp Flix, Switzerland, O3 fumigation
(AOT40 of 15.2-64.9 ppm-h) showed no significant impact on community
composition of this species-rich pasture (Bassin et al., 2007b). Although most studies
demonstrated an increase in grass:forb ratio with O3 exposure (Haves et al., 2009;
U.S. EPA, 2006b), a study on a simulated upland grassland community showed that
O3 reduced the grass:forb ratio (Haves et al.. 2010) which may be due to the grass
species in this community. The grass species studied by Hayes et al. (2010).
Anthoxanthum odoratum, was more sensitive to O3 than other grass species such as
L. perenne (Haves et al.. 2009). Pfleeger et al. (2010) collected seed bank soil from
an agricultural field and examined how the plant community responded over several
generations to elevated O3 exposures. Sixty plant species from 22 families emerged
in the chambers over their four year study. Overall, they found that O3 appeared to
have small impacts on seed germination and only a minor effect on species richness
of pioneer plant communities.
Several review papers have discussed the physiological and ecological characteristics
of O3-sensitive herbaceous plants. Hayes et al. (2007) assessed species traits
associated with O3 sensitivity by the changes in biomass caused by O3 exposure.
Plants of the therophyte (e.g., annual) life form were particularly sensitive to O3.
Species with higher mature leaf N concentration tended to be more sensitive than
those with lower leaf N concentration. Plants growing under high oxidative stress
environments, such as high light or high saline, were more sensitive to O3. Using the
same dataset from Hayes et al. (2007), Mills et al. (2007b) identified the O3 sensitive
communities. They found that the largest number of these O3 sensitive communities
were associated with grassland ecosystems. Among grassland ecosystems, alpine
grassland, sub-alpine grassland, woodland fringe, and dry grassland were identified
as the most sensitive communities.
9-82
-------
9.4.7.3 Microbes
Several methods have been used to study microbial composition changes associated
with elevated O3. Phospholipid fatty acid (PLFA) analysis is widely used to
determine whether O3 elicits an overall effect on microbial community composition.
However, since PLFA markers cover a broad range of different fungi, resolution of
this method may be not fine enough to detect small changes in the composition of
fungal communities. Methods, such as microscopic analyses and polymerase chain
reaction-denaturing gradient gel electrophoresis (PCR-DGGE), have better
resolution to specifically analyze the fungal community composition. The resolution
differences among those methods needs to be considered when assessing the O3
impact on microbial community composition.
Kanerva et al. (2008) found that elevated O3 (40-50 ppb) decreased total, bacterial,
actinobacterial and fungal PLFA biomass values as well as fungal:bacterial PLFA
biomass ratio in their meadow mesocosms in south-western Finland. The relative
proportions of individual PLFAs between the control and elevated O3 treatments
were significantly different, suggesting that O3 modified the structure of the
microbial community. Morsky et al. (2008) exposed boreal peatland microcosms to
elevated O3, with growing season AOT40 of 20.8-35.3 ppm-h, in an open-air O3
exposure field in Central Finland. They also found that microbial composition was
altered after three growing seasons with O3 fumigation, as measured by PLFA.
Ozone tended to increase the presence of Gram-positive bacteria and the biomass of
fungi in the peatland microcosms. Ozone also resulted in higher microbial biomass,
which co-occurred with the increases in concentrations of organic acids and leaf
density of sedges (Morsky et al., 2008). In a lysimeter experiment in Germany, O3
was found to alter the PLFA profiles in the upper 0-20 cm rhizosphere soil of
European beech. Elevated O3 reduced bacterial abundance but had no detectable
effect on fungal abundance (Pritsch et al., 2009). Using microscopic analyses,
Kasurinen et al. (2005) found that elevated O3, with 5 or 6 months of AOT40 of
20.6-30.9 ppm-h, decreased the proportions of black and liver-brown mycorrhizas
and increased that of light brown/orange mycorrhizas. In an herbaceous plant study,
SSCP (single-strand conformation polymorphism) profiles indicated that O3 stress
(about 75 ppb) had a very small effect on the structural diversity of the bacterial
community in rhizospheres (Dohrmann and Tebbe, 2005). At the Aspen FACE site,
O3 had no significant effect on fungal relative abundance, as indicated by PLFA
profile. However, elevated O3 altered fungal community composition, according to
the identification of 39 fungal taxonomic units from soil using polymerase chain
reaction-denaturing gradient gel electrophoresis (PCR-DGGE) (Chung et al., 2006).
In another study at Aspen FACE, phylogenetic analysis suggested that O3 exposure
altered the agaricomycete community. The ectomycorrhizal communities developing
under elevated O3 had higher proportions of Cortinarius and Inocybe species, and
lower proportions ofLaccaria and Tomentella (Edwards and Zak, 2011). Ozone was
found to change microbial community composition in an agricultural system. Chen et
al. (201 Ob) found elevated O3 (100-150 ppb) had significant effects on soil microbial
composition expressed as PLFA percentage in a rice paddy in China.
9-83
-------
9.4.7.4 Summary
In the 2006 O3 AQCD, the impact of O3 exposure on species competition and
community composition was assessed. Ozone was found to cause a significant
decline in ponderosa and Jeffrey pine in the San Bernardino Mountains in southern
California. Ozone exposure also tended to shift the grass-legume mixtures in favor of
grass species (U.S. EPA. 2006b). Since the 2006 O3 AQCD, more evidence has
shown that O3 exposure changed the competitive interactions and could lead to loss
of O3 sensitive species or genotypes. Studies at plant level found that the severity of
O3 damage on growth, reproduction, and foliar injury varied among species, which
provided the biological plausibility for the alteration of community composition.
Additionally, research since the last review has shown that O3 can alter community
composition and diversity of soil microbial communities.
The decline of conifer forests under O3 exposure was continually observed in several
regions. Ozone damage was believed to be an important causal factor in the dramatic
decline of sacred fir in the valley of Mexico (de Lourdes de Bauer and Hernandez-
Tejeda. 2007). as well as cembran pine in southern France and the Carpathian
Mountains (Wieser et al.. 2006). Results from the Aspen FACE site indicated that O3
could alter community composition of broadleaf forests as well. At the Aspen FACE
site, O3 reduced growth and increased mortality of a sensitive aspen clone, while the
O3 tolerant clone emerged as the dominant clone in the pure aspen community. In the
mixed aspen-birch and aspen-maple communities, O3 reduced the competitive
capacity of aspen compared to birch and maple (Kubiske et al.. 2007).
The tendency for O3-exposure to shift the biomass of grass-legume mixtures in favor
of grass species, was reported in the 2006 O3 AQCD and has been generally
confirmed by recent studies. However, in a high elevation mature/species-rich grass-
legume pasture, O3 fumigation showed no significant impact on community
composition (Bassin et al.. 2007b).
Ozone exposure not only altered community composition of plant species, but also
microorganisms. The shift in community composition of bacteria and fungi has been
observed in both natural and agricultural ecosystems, although no general patterns
could be identified (Kanerva et al., 2008; Morsky et al., 2008; Kasurinen et al.,
2005).
The evidence is sufficient to conclude that there is likely to be a causal
relationship between O3 exposure and the alteration of community
composition of some ecosystems.
9.4.8 Factors that Modify Functional and Growth Response
Many biotic and abiotic factors, including insects, pathogens, root microbes and
fungi, temperature, water and nutrient availability, and other air pollutants, as well as
elevated CO2, influence or alter plant response to O3. These modifying factors were
9-84
-------
comprehensively reviewed in AX9.3 of the 2006 O3 AQCD and thus, this section
serves mainly as a brief summary of the previous findings. A limited number of new
studies published since the 2006 O3 AQCD add to the understanding of the role of
these interactions in modifying O3-induced plant responses. Many of these
modifying factors and interactions are integrated into discussions elsewhere in this
chapter and the reader is directed to those sections.
9.4.8.1 Genetics
It is well known that species vary greatly in their responsiveness to O3. Even within a
given species, individual genotypes or populations can also vary significantly with
respect to O3 sensitivity (U.S. EPA, 2006b). Therefore, caution should be taken when
considering a species' degree of sensitivity to O3. Plant response to O3 is determined
by genes that are directly related to oxidant stress and to an unknown number of
genes that are not specifically related to oxidants, but instead control leaf and cell
wall thickness, stomatal conductance, and the internal architecture of the air spaces.
It is rarely the case that single genes are responsible for O3 tolerance. Studies using
molecular biological tools and transgenic plants have positively verified the role of
various genes and gene products in O3 tolerance and are continuing to increase the
understanding of O3 toxicity and differences in O3 sensitivity. See Section 9.3.3.2 of
this document for a discussion of recent studies related to gene expression changes in
response to O3.
9.4.8.2 Environmental Biological Factors
As stated in the 2006 O3 AQCD, the biological factors within the plant's
environment that may influence its response to O3 encompass insects and other
animal pests, diseases, weeds, and other competing plant species. Ozone may
influence the severity of a disease or infestation by a pest or weed, either by direct
effects on the causal species, or indirectly by affecting the host, or both. In addition,
the interaction between O3, a plant, and a pest, pathogen, or weed may influence the
response of the target host species to O3 (U.S. EPA, 2006b). Several recent studies
on the effects of O3 on insects via their interactions with plants are discussed in
Section 9.4.9.1 In addition, O3 has also been shown to alter soil fauna communities
(Section 9.4.9.2).
In contrast to detrimental biological interactions, there are mutually beneficial
relationships or symbioses involving higher plants and bacteria or fungi. These
include (1) the nitrogen-fixing species Rhizobium andFrankia that nodulate the roots
of legumes and alder and (2) the mycorrhizae that infect the roots of many crop and
tree species, all of which may be affected by exposure of the host plants to O3. Some
discussion of mycorrhizae can be found in Section 9.4.6.
9-85
-------
In addition to the interactions involving animal pests, O3 also has indirect effects on
higher herbivorous animals, e.g., livestock, due to O3-induced changes in feed
quality. Recent studies on the effects of O3 on nutritive quality of plants are
discussed in Section 9.4.4.2.
Intra- and interspecific competition are also important factors in determining
vegetation response to O3. Plant competition involves the ability of individual plants
to acquire the environmental resources needed for growth and development: light,
water, nutrients, and space. Intraspecific competition involves individuals of the
same species, typically in monoculture crop situations, while interspecific
competition refers to the interference exerted by individuals of different species on
each other when they are in a mixed culture. This topic was previously reviewed in
AX9.3.3.4 of the 2006 O3 AQCD. Recent studies on competition and its implications
for community composition are discussed in Section 9.4.7.
9.4.8.3 Physical Factors
Physical or abiotic factors play a large role in modifying plant response to O3, and
have been extensively discussed in previous O3 AQCDs. This section summarizes
those findings as well as recent studies published since the last review.
Although some studies have indicated that O3 impact significantly increases with
increased ambient temperature (Ball et al., 2000; Mills et al., 2000), other studies
have indicated that temperature has little effect (Balls et al., 1996; Fredericksen et al.,
1996). A recent study by Riikonen et al. (2009) at the Ruohoniemi open air exposure
field in Kuopio, Finland found that the effects of temperature and O3 on total leaf
area and photosynthesis of Betulapendula were counteractive. Elevated O3 reduced
the saplings' ability to utilize the warmer growth environment by increasing the
stomatal limitation for photosynthesis and by reducing the redox state of ascorbate in
the apoplast in the combination treatment as compared to temperature alone
(Riikonen et al.. 2009).
Temperature affects the rates of all physiological processes based on enzyme
catalysis and diffusion; each process and overall growth (the integral of all processes)
has a distinct optimal temperature range. It is important to note that a plant's
response to changes in temperature will depend on whether it is growing near its
optimum temperature for growth or near its maximum temperature (Rowland-
Bamford, 2000). However, temperature is very likely an important variable affecting
plant O3 response in the presence of the elevated CO2 levels contributing to global
climate change. In contrast, some evidence suggests that O3 exposure sensitizes
plants to low temperature stress (Colls andUnsworth, 1992) and, also, that O3
decreases below-ground carbohydrate reserves, which may lead to responses in
perennial species ranging from rapid demise to impaired growth in subsequent
seasons (i.e., carry-over effects) (Andersen et al., 1997).
9-86
-------
Light, a component of the plant's physical environment, is an essential "resource" of
energy content that drives photosynthesis and C assimilation. It has been suggested
that increased light intensity may increase the O3 sensitivity of light-tolerant species
while decreasing that of shade-tolerant species, but this appears to be an
oversimplification with many exceptions. Several studies suggest that the interaction
between O3 sensitivity and light environment is complicated by the developmental
stage as well as the light environment of individual leaves in the canopy (Kitao et al..
2009: Topaetal..2Q01: Chappelka and Samuelson. 1998).
Although the relative humidity of the ambient air has generally been found to
increase the effects of O3 by increasing stomatal conductance (thereby increasing O3
flux into the leaves), abundant evidence also indicates that the ready availability of
soil moisture results in greater O3 sensitivity (Mills, 2002). The partial "protection"
against the effects of O3 afforded by drought has been observed in field experiments
(Low et al., 2006) and modeled in computer simulations (Broadmeadow and Jackson,
2000). Conversely, drought may exacerbate the effects of O3 on plants (Pollastrini et
al., 2010; Grulke et al., 2003b). There is also some evidence that O3 can predispose
plants to drought stress (Maier-Maercker, 1998). Hence, the nature of the response is
largely species-specific and will depend to some extent upon the sequence in which
the stressors occur.
9.4.8.4 Interactions with Other Pollutants
Ozone-nitrogen interactions
Elevated O3 exposure and N deposition often co-occur. However, the interactions of
O3 exposure and N deposition on vegetation are complex and less well understood
compared to their independent effects. Consistent with the conclusion of the 2006 O3
AQCD, the limited number of studies published since the last review indicated that
the interactive effects of N and O3 varied among species and ecosystems (Table 9-8).
Nitrogen deposition could stimulate relative growth rate (RGR), and lead to
increased stomatal conductance. Therefore, plants might become more susceptible to
O3 exposure. Alternatively, N deposition may increase the availability of
photosynthates for use in detoxification and plants could become more tolerant to O3
(Bassin et al.. 2007a). Elevated O3 exposure and N deposition could also act in
concert to increase plant susceptibility to disease (von Tiedemann. 1996). To better
understand these interactions in ecosystems across the U.S., more information is
needed considering combined O3 exposure and N deposition related effects.
Only a few recent studies have investigated the interactive effects of O3 and N in the
United States. Grulke et al. (2005) measured stomatal conductance of California
black oak (Quercus kelloggii) at a long-term N-enrichment site located in the San
Bernardino Mountains, which is accompanied by high O3 exposure (80 ppb,
24-h avg. over a six month growing season). The authors found that N amendment
led to poor stomatal control in full sun in midsummer of the average precipitation
9-87
-------
years, but enhanced stomatal control in shade leaves of California black oak. In an
OTC study, Handley and Grulke (2008) found that O3 lowered photosynthetic ability
and water-use efficiency, and increased leaf chlorosis and necrosis of California
black oak. Nitrogen fertilization tended to reduce plant sensitivity to O3 exposure;
however, the interaction was not statistically significant. In another study, Grulke et
al. (2008) reported that various lines of phenomenological and experimental evidence
indicate that N deposition and O3 pollution contribute to the susceptibility of forests
to wildfire in the San Bernadino Mountains by increasing stress due to drought,
weakening trees, and predisposing them to bark beetle infestation (U.S. EPA. 2008b).
Studies conducted outside the U.S. are also summarized in Table 9-8. Generally, the
responses were species specific. The O3-induced reduction in photosynthetic rate and
biomass loss were greater in the relatively high N treatment for watermelon (Citrillus
lanants) (Calatayud et al., 2006) and Japanese beech (Fagus crenata) seedlings
(Yamaguchi et al., 2007). However, there was no significant interactive effect of O3
and N on biomass production for Quercus serrata seedlings (Watanabe et al., 2007),
young Norway spruce (Picea abies) trees (Thomas et al., 2005), and young European
beech (Fagus sylvatica) trees Thomas et al. (2006).
-------
Table 9-8 Response of plants to the interactive effects of elevated O3
exposure and nitrogen enrichment.
Site
San
Bernardino
Mountains,
U.S.
San
Bernardino
Mountains,
U.S.
Switzerland
Switzerland
Switzerland
Switzerland
Switzerland
Species
California
black oak
(Quercus
kelloggii)
California
black oak
(Quercus
kelloggii)
Spruce trees
(Picea
abies)
Beech trees
(Fagus
sylvatica)
Alpine
pasture
Alpine
pasture
Alpine
pasture
Ozone exposure
80 ppb
0, 75, and 150 ppb
Filtered
(19.4-28.1 ppb);
Ambient
(37.6-47.4 ppb)
Filtered
(19.4-28.1 ppb);
Ambient
(37.6-47.4 ppb)
Ambient
(AOT40 of
11. 1-12.6 ppm-h);
1 .2 ambient
(AOT40 of
15.2-29.5 ppm-h);
1.6 ambient
(28.4-64.9 ppm-h)
Ambient
(AOT40 of
11. 1-12.6 ppm-h);
1 .2 ambient
(AOT40 of
15.2-29.5 ppm-h);
1 .6 ambient
(28.4-64.9 ppm-h)
Ambient
(AOT40 of
11. 1-12.6 ppm-h);
1.2 ambient
(AOT40 of
15.2-29.5 ppm-h);
1.6 ambient
(28.4-64.9 ppm-h)
N addition
0, and
50 kg N/ha-yr
0, and
50 kg N/ha-yr
0, 20, 40 and 80
kg N/ha-yr
0, 20, 40 and 80
kg N/ha-yr
0, 5, 10,25, 50
kg N/ha-yr
0,5, 10,25,50
kg N/ha-yr
0, 5, 10' 25, 50
kg N/ha-yr
Responses
N-amended trees had lower late
summer C gain and greater foliar
chlorosis in the drought year, and
poor stomatal control and lower
leaf water use efficiency and in
midsummer of the average
precipitation year.
N fertilization tended to reduce
plant sensitivity to O3 exposure;
however the interaction was not
statistically significant.
Higher N levels alleviated the
negative impact of O3 on root
starch concentrations
N addition amplified the negative
effects of O3 on leaf area and
shoot elongation.
The positive effects of N addition
on canopy greenness were
counteracted by accelerated leaf
senescence in the highest O3
treatment.
Only a small number of species
showed significant O3 and N
interactive effects on leaf
chlorophyll concentration, leaf
weight and change in 18O, and the
patterns were not consistent.
Highest N addition resulted in
carbon loss, but there was no
interaction between O3 and
N treatments.
References
Grulke et al.
(2005)
Handley and
Grulke
(2008)
Thomas et al.
(2005)
Thomas et al.
(2006)
Bassin et al.
(2007b)
Bassin et al.
(2009)
Volketal.
(2011)
9-89
-------
Site
Spain
Spain
Japan
Japan
Species
Watermelon
(Citrillus
lanants)
Clover
Trifolium
striatum
Japanese
beech
seedlings
(Fagus
crenata)
Japanese
tree
(Quercus
serrata)
seedlings
Ozone exposure N addition
O3free 140, 280, and
(AOT40 of 436 kg N/ha-yr
0 ppm-h),
Ambient
(AOT40 of
5.1 -6.3 ppm-h);
Elevated O3
(AOT40 of
32.5-35.6 ppm-h)
Filtered 10, 30, and 60
(24-h avg. of kg N/ha-yr
8-22 ppb);
Ambient
(29-34 ppb),
Elevated O3
(35-56 ppb)
Filtered 0, 20 and 50
(24-h avg. of kg N/ha-yr
10.3-1 3.2 ppb);
Ambient
(42.0-43.3 ppb),
1.5 Ambient
(62.6-63.9 ppb);
2.0 ambient
(82.7-84.7 ppb)
Filtered 0, 20 and 50
(24-h avg. of kg N/ha-yr
10.3-1 3.2 ppb);
Ambient
(42.0-43.3 ppb),
1.5 ambient
(62.6-63.9 ppb);
2.0 ambient
(82.7-84.7 ppb)
Responses
High N concentration enhanced
the detrimental effects of O3 on
Chlorophyll a fluorescence
parameters, lipid peroxidation,
and the total yield.
O3 reduced total aerial biomass.
N fertilization counterbalanced
O3-induced effects only when
plants were exposed to moderate
O3 levels (ambient) but not under
elevated O3 concentrations.
The Os-induced reduction in net
photosynthesis and whole-plant
dry mass were greater in the
relatively high N treatment than
that in the low N treatment.
No significant interactive effects
of O3 and N load on the growth
and net photosynthetic rate were
detected.
References
Calatayud et al.
(2006)
Sanz et al.
(2007)
Yamaguchi et al.
(2007)
Watanabe et.al.
(2007)
Ozone-carbon dioxide interactions
Several decades of research has shown that exposure to elevated CO2 increases
photosynthetic rates (Bernacchi et al., 2006; Bernacchi et al., 2005; Tissue et al.,
1999; Tissue et al., 1997; Will and Ceulemans, 1997), decreases stomatal
conductance (Ainsworth and Rogers, 2007; Paoletti et al., 2007; Bernacchi et al.,
2006; Leakey et al., 2006; Medlyn et al., 2001) and generally increases the growth of
plants (McCarthy et al., 2009; Norby et al., 2005). This is in contrast to the decrease
in photosynthesis and growth in many plants that are exposed to elevated O3.
The interactive effects on vegetation have been the subject of research in the past two
decades due to the implications on productivity and water use of ecosystems. This
area of research was discussed in detail in AX9.3.8.1 of the 2006 O3 AQCD and the
conclusions made then are still relevant (U.S. EPA, 2006b).
9-90
-------
The bulk of the available evidence shows that, under the various experimental
conditions used (which almost exclusively employed abrupt or "step" increases in
CO2 concentration, as discussed below), increased CO2 levels (ambient + 200 to
400 ppm) may protect plants from the negative effects of O3 on growth. This
protection may be afforded in part by CO2 acting together with O3 in inducing
stomatal closure, thereby reducing O3 uptake, and in part by CO2 reducing the
negative effects of O3 on Rubisco and its activity in CO 2-fixation. Although both
CO2-induced and O3-induced decreases in stomatal conductance have been observed
primarily in short-term studies, recent data show a long-term and sustained reduction
in stomatal conductance under elevated CO2 for a number of species (Ainsworth and
Long. 2005: Ellsworth et al.. 2004: Gunderson et al.. 2002). Instances of increased
stomatal conductance have also been observed in response to O3 exposure,
suggesting partial stomatal dysfunction after extended periods of exposure (Paoletti
and Grulke. 2010: Grulke et al.. 2007a: Maier-Maercker. 1998).
Important caveats must be raised with regard to the findings presented in published
research. The first caveat concerns the distinctly different natures of the exposures to
O3 and CO2 experienced by plants in the field. Changes in the ambient
concentrations of these gases have very different dynamics. In the context of climate
change, CO2 levels increase relatively slowly (globally 2 ppm/year) and may change
little over several seasons of growth. On the other hand, O3 presents a fluctuating
stressor with considerable hour-to-hour, day-to-day and regional variability (Polle
and Pell, 1999). Almost all of the evidence presented comes from experimentation
involving plants subjected to an abrupt step increase to a higher, steady CO2
concentration. In contrast, the O3 exposure concentrations usually varied from day to
day. Luo and Reynolds (1999), Hui et al. (2002), and Luo (2001) noted the
difficulties in predicting the likely effects of a gradual CO2 increase from
experiments involving a step increase or those using a range of CO2 concentrations.
It is also important to note that the levels of elevated CO2 in many of the studies will
not be experienced in the field for 30 or 40 years, but elevated levels of O3 can occur
presently in several areas of the United States. Therefore, the CO2 x O3 interaction
studies may be less relevant for current ambient conditions.
Another caveat concerns the interactions of O3 and CO2 with other climatic
variables, such as temperature and precipitation. In light of the key role played by
temperature in regulating physiological processes and modifying plant response to
increased CO2 levels (Morison and Lawlor, 1999: Long, 1991) and the knowledge
that relatively modest increases in temperature may lead to dramatic consequences in
terms of plant development (Lawlor, 1998), it is important to consider that studying
CO2 and O3 interactions alone may not create a complete understanding of effects on
plants under future climate change.
9-91
-------
9.4.9 Insects and Other Wildlife
9.4.9.1 Insects
Insects may respond indirectly to changes in plants (i.e., increased reactive oxygen
species, altered phytochemistry, altered nutrient content) that occur under elevated
O3 conditions, or O3 can have a direct effect on insect performance (Menendez et al.,
2009). Effects of O3 on insects occur at the species level (i.e., growth, survival,
reproduction, development, feeding behavior) and at the population and community-
level (i.e., population growth rate, community composition). In general, effects of O3
on insects are highly context- and species-specific (Lindroth, 2010; Bidart-Bouzat
and Imeh-Nathaniel, 2008). Furthermore, plant responses to O3 exposure and
herbivore attack have been demonstrated to share signaling pathways, complicating
characterization of these stressors (Lindroth, 2010; Menendez et al., 2010, 2009).
Although both species-level and population and community-level responses to
elevated O3 are observed in field and laboratory studies discussed below, there is no
consensus on how insects respond to feeding on O3-exposed plants.
Species-level responses
In considering insect growth, survival and reproduction in elevated O3 conditions,
several studies have indicated an effect while others have found no correlation.
The performance of five herbivore species (three moths and two weevils) was
assessed in an OTC experiment at 2 x ambient concentration (Peltonen et al., 2010).
Growth of larvae of the Autumnal moth, Epirrita autumna, was significantly
decreased in the O3 treatment while no effects were observed in the other species.
In an aphid oviposition preference study using birch buds grown in a three year OTC
experiment, O3 had neither a stimulatory or deterring effect on egg-laying (Peltonen
et al., 2006). Furthermore, changes in birch bud phenolic compounds associated with
the doubled ambient concentrations of O3 did not correlate with changes in aphid
oviposition (Peltonen et al., 2006). Reproduction in Popilliajaponica, that were fed
soybeans and grown under elevated O3 appeared to be unaffected (O'Neill et al.,
2008). In a meta-analysis of effects of elevated O3 on 22 species of trees and 10
species of insects, the rates of survival, reproduction and food consumption were
typically unaffected while development times were reduced and pupal masses were
increased (Valkama et al.. 2007).
At the Aspen FACE site insect performance under elevated (50-60 ppb) O3
conditions (approximately 1.5 x background ambient levels of 30-40 ppb O3) have
been considered for several species. Cumulative fecundity of aphids (Cepegillettea
betulaefoliae), that were reared on O3-exposed paper birch (Betulapapyri/era) trees,
was lower than aphids from control plots (Awmack et al., 2004). No effects on
growth, development, adult weight, embryo number and birth weight of newborn
nymphs were observed. In a study conducted using three aspen genotypes,
9-92
-------
performance of the aspen beetle (Chrysomela crochi) decreased across all parameters
measured (development time, adult mass and survivorship) under elevated O3 (Vigue
and Lindroth, 2010). There was an increase in the development time of male and
female aspen beetle larvae although the percentages varied across genotypes.
Decreased beetle adult mass and survivorship was observed across all genotypes
under elevated O3 conditions. Another study from the Aspen FACE site did not find
any significant effects of elevated O3 on performance (longevity, fecundity,
abundance) of the invasive weevil (Polydrusus sericeus) (Hillstrom et al.. 201 Ob).
Since the 2006 O3 AQCD, several studies have considered the effect of elevated O3
on feeding behavior of insects. In a feeding preference study, the common leaf
weevil (Phyllobius pyri) consumed significantly more leaf discs from one aspen
clone when compared to a second clone under ambient air conditions (Freiwald et al.,
2008). In a moderately elevated O3 environment (1.5 x ambient), this preference for
a certain aspen clone was less evident, however, leaves from O3-exposed trees were
significantly preferred to leaves grown under ambient conditions. Soybeans grown
under enriched O3 had significantly less loss of leaf tissue to herbivory in August
compared to earlier in the growing season (July) when herbivory was not affected
(Hamilton et al., 2005). Other plant-herbivore interactions have shown no effects of
elevated O3 on feeding. Feeding behavior of Japanese beetles (P. japonicd) appeared
to be unchanged when beetles were fed soybean leaves grown under elevated O3
conditions (O'Neill et al., 2008). At the Aspen FACE site, feeding by the invasive
weevil (Polydrusus sericeus), as measured by leaf area consumption, was not
significantly different between foliage that was grown under elevated O3 versus
ambient conditions (Hillstrom et al., 201 Ob).
Population-level and community-level responses
Recent data on insects provide evidence of population-level and community-level
responses to O3. Elevated levels of O3 can affect plant phytochemistry and nutrient
content which in turn can alter population density and structure of the associated
herbivorous insect communities and impact ecosystem processes (Cornelissen, 2011;
Lindroth, 2010). In 72-hour exposures to elevated O3, mean relative growth rate of
the aphid Diuraphis noxia increased with O3 concentration suggesting that more
rapid population growth may occur when atmospheric O3 is elevated (Summers et
al., 1994). In a long-term study of elevated O3 on herbivore performance at the
Aspen FACE site, individual performance and population-level effects of the aphid
C. betulaefoliae were assessed. Elevated O3 levels had a strong positive effect on the
population growth rates of the aphids; although effects were not detected by
measuring growth, development, adult weight, embryo number or birth weight of
newborn nymphs (Awmack et al., 2004). Conversely, a lower rate of population
growth was observed in aphids previously exposed to O3 in an OTC (Menendez et
al.. 2010). No direct effects of O3 were observed; however, nymphs born from adults
exposed to and feeding on O3 exposed plants were less capable of infesting new
plants when compared to nymphs in the control plots (Menendez et al.. 2010).
Elevated O3 reduced total arthropod abundance by 17% at Aspen FACE, largely as a
9-93
-------
result of the negative effects on parasitoids, although phloem-feeding insects may
benefit (Hillstrom and Lindroth, 2008). Herbivore communities affected by O3 and N
were sampled along an air pollution gradient in the Los Angeles basin (Jones and
Paine, 2006). Abundance, diversity, and richness of herbivores were not affected.
However, a shift in community structure, from phloem-feeding to chewing
dominated communities, was observed along the gradient. No consistent effect of
elevated O3 on herbivory or insect population size was detected at SoyFACE
(O'Neill et al.. 2010: Dermodv et al.. 2008).
Evidence of modification of insect populations and communities in response to
elevated O3 includes genotypic and phenotypic changes. In a study conducted at the
Aspen FACE site, elevated O3 altered the genotype frequencies of the pea aphid
(Acyrtho siphon pi sum) grown on red clover (Trifolium pratense) over multiple
generations (Mondor et al., 2005). Aphid color was used to distinguish between the
two genotypes. Ozone increased the genotypic frequencies of
pink-morph:green-morph aphids from 2:1 to 9:1, and depressed wing-induction
responses more strongly in the pink than the green genotype (Mondor et al., 2005).
Growth and development of individual green and pink aphids reared as a single
genotype or mixed genotypes were unaffected by elevated O3 (Mondor et al., 2010).
However, growth of pea aphid populations is not readily predictable using individual
growth rates.
9.4.9.2 Wildlife
Herpetofauna
Since the 2006 O3 AQCD, direct effects of O3 exposure including physiological
changes and alterations of ecologically important behaviors such as feeding and
thermoregulation have been observed in wildlife. These studies have been conducted
in limited laboratory exposures, and the levels of O3 treatment (e.g., 0.2-0.8 ppm)
were often unrealistically higher than the ambient levels. Amphibians may be
especially vulnerable to airborne oxidants due to the significant gas exchange that
occurs across the skin (Andrews et al., 2008; Dohm et al., 2008). Exposure to
0.2 ppm to 0.8 ppm O3 for 4 hours resulted in a decrease of oxygen consumption and
depressed lung ventilation in the California tree frog Pseudacris cadaverina (Mautz
and Dohm. 2004). Following a single 4-h inhalation exposure to 0.8 ppm O3, reduced
pulmonary macrophage phagocytosis was observed at 1 and 24 hours postexposure
in the marine toad (Bufo marinus) indicating an effect on immune system function
(Dohm et al.. 2005). There was no difference in macrophage function at 48 hours
postexposure in exposed and control individuals.
Behavioral effects of O3 observed in amphibians include responses to minimize the
surface area of the body exposed to the air and a decrease in feeding rates (Dohm et
al., 2008; Mautz and Dohm, 2004). The adoption of a low-profile "water
conservation posture" during O3 exposure was observed in experiments with the
9-94
-------
California tree frog (Mautz and Dohm, 2004). Marine toads, Bufo marinus, exposed
to 0.06 j^L/L (ppm) O3 for 4 hours ate significantly fewer mealworms at 1 hour and
48 hours postexposure than control toads (Dohm et al., 2008). In the same study,
escape/exploratory behavior as measured by total distance moved was not negatively
affected in the O3-exposed individuals as compared to the controls (Dohm et al.,
2008).
Water balance and thermal preference in herpetofauna are altered with elevated O3.
Marine toads exposed to 0.8 ppm O3 for 4 hours exhibited behavioral hypothermia
when temperature selection in the toads was assessed at 1, 24 and 48 hours
postexposure (Dohm et al., 2001). Ozone-exposed individuals lost almost 5g more
body mass on average than controls due to evaporative water loss. At 24 hours after
exposure, the individuals that had lost significant body mass selected lower body
temperatures (Dohm et al., 2001). Behavioral hypothermia was also observed in
reptiles following 4-h exposures to 0.6 ppm O3. Exposure of the Western Fence
Lizard (Sceloporus occidentalis) at 25°C induced behavioral hypothermia that
recovered to control temperatures by 24 hours (Mautz and Dohm, 2004).
The behavioral hypothermic response persisted in lizards exposed to O3 at 35°C at
24 hours postexposure resulting in a mean body temperature of 3.3°C over controls.
Soil fauna communities
Ozone has also been shown to alter soil fauna communities (Meehan et al., 2010;
Kasurinen et al., 2007; Loranger et al., 2004). Abundance of Acari (mites and ticks)
decreased by 47% under elevated O3 at Aspen FACE site, probably due to the higher
secondary metabolites and lower N concentrations in litter and foliage under elevated
O3 (Loranger et al., 2004). In another study from the Aspen FACE site, leaf litter
collected from aspen grown under elevated O3 conditions was higher in fiber and
lignin concentrations than litter from trees grown under ambient conditions. These
chemical characteristics of the leaves were associated with increased springtail
population growth following 10 weeks in a laboratory microcosm (Meehan et al.,
2010). Consumption rates of earthworms fed on leaf litter for 6 weeks from trees
grown under elevated O3 conditions and ambient air did not vary significantly
between treatments (Meehan et al., 2010). In another study on juvenile earthworms
Lumbricus terrestris, individual growth was reduced when worms were fed high-O3
birch litter from trees exposed for three years to elevated O3 in an OTC system
(Kasurinen et al., 2007). In the same study no significant growth or mortality effects
were observed in isopods.
9.4.9.3 Indirect Effects on Wildlife
In addition to the direct effects of O3 exposure on physiological and behavioral
endpoints observed in the laboratory, there are indirect effects to wildlife. These
effects include changes in biomass and nutritive quality of O3-exposed plants
(reviewed in Section 9.4.4) that are consumed by wildlife. Reduced digestibility of
9-95
-------
O3-exposed plants may alter dietary intake and foraging strategies in herbivores. In a
study using native highbush blackberry (Rubus argutus) relative feed value of the
plants decreased in bushes exposed to double ambient concentrations of O3
(Ditchkoff et al., 2009). Indirect effects of elevated O3 on wildlife include changes in
chemical signaling important in ecological interactions reviewed below.
Chemical signaling in ecological interactions
Ozone has been shown to degrade or alter biogenic VOC signals important to
ecological interactions including; (1) attraction of pollinators and seed dispersers; (2)
defense against herbivory; and (3) predator-prey interactions (Pinto et al., 2010;
McFrederick et al.. 2009: Yuan et al.. 2009: Pinto et al.. 2007a: Pinto et al.. 2007b).
Each signal released by emitters has an atmospheric lifetime and a unique chemical
signature comprised of different ratios of individual hydrocarbons that are
susceptible to atmospheric oxidants such as O3 (Yuan et al., 2009; Wright et al.,
2005). Under elevated O3 conditions, these olfactory cues may travel shorter
distances before losing their specificity (McFrederick et al., 2009; McFrederick et al.,
2008). Additional non-phytogenic VOC-mediated interrelationships with the
potential to be modified by O3 include territorial marking, pheromones for attraction
of mates and various social interactions including scent trails, nestmate recognition
and signals involved in aggregation behaviors (McFrederick et al., 2009). For
example, the alcohols, ketones and aldehydes comprising sex pheromones in moths
could be especially vulnerable to degradation by O3, since some males travel >100
meters to find mates (Garde and Havnes, 2004). In general, effects of O3 on scent-
mediated ecological interactions are highly context- and species-specific (Lindroth,
2010: Bidart-Bouzat and Imeh-Nathaniel. 2008).
Pollination and seed dispersal
Phytogenic VOC's attract pollinators and seed dispersers to flowers and fruits
(Dudareva et al., 2006; Theis and Raguso, 2005). These floral scent trails in
plant-insect interactions may be destroyed or transformed by O3 (McFrederick et al.,
2008). Using aLagrangian model, the rate of destruction of phytogenic VOC's was
estimated in air parcels at increasing distance from a source in response to increased
regional levels of O3, hydroxyl and nitrate radicals (McFrederick et al., 2008). Based
on the model, the ability of pollinators to locate highly reactive VOCs from emitting
flowers may have decreased from kilometers during pre-industrial times to <200
meters at current ambient conditions (McFrederick et al., 2008). Scents that travel
shorter distances (0-10 meters) are less susceptible to air pollutants, while highly
reactive scents that travel longer distances (10 to 100s of meters), are at a higher risk
for degradation (McFrederick et al., 2009). For example, male euglossine bees can
detect bait stations from a distance of at least one kilometer (Dobson, 1994).
9-96
-------
Defense against herbivory
Ozone can alter the chemical signature of VOCs emitted by plants and these VOCs
are subsequently detected by herbivores (Blande et al., 2010; Iriti and Faoro, 2009;
Pinto et al.. 2007a; Vuorinen et al.. 2004; Jackson et al.. 1999; Cannon. 1990). These
modifications can make the plant either more attractive or repellant to phytophagous
insects (Pinto et al., 2010). For example, under elevated O3, the host plant preference
by forest tent caterpillars increased for birch compared to aspen (Agrell et al., 2005).
Ozone-induced emissions from red spruce needles were found to repel spruce
budworm larvae (Cannon, 1990). Transcriptional profiles of field grown soybean
(Glycine max) grown in elevated O3 conditions were altered due to herbivory by
Japanese beetles. The herbivory resulted in a higher number of transcripts in the
leaves of O3-exposed plants and upregulation of antioxidant metabolism associated
with plant defense (Casteel et al., 2008).
Ozone may modify signals involved in plant-to-plant interactions and plant defense
against pathogens (Blande et al.. 2010; Pinto et al.. 2010; McFrederick et al.. 2009;
Yuan et al.. 2009). In a recent study with lima beans, 80 ppb O3 degraded several
herbivore-induced VOCs, reducing the distance over which plant-to-plant signaling
occurred (Blande et al.. 2010).
Predator-prey interactions
Elevated O3 conditions are associated with disruption of pheromone-mediated
interactions at higher trophic levels (e.g., predators and parasitoids of herbivores).
In a study from the Aspen FACE site, predator escape behaviors of the aphid
(Chatophorus stevensis) were enhanced on O3-fumigated aspen trees although the
mechanism of this response remains unknown (Mondor et al.. 2004). The predatory
mite Phytoseiulus persimilis can distinguish between the VOC signature of ozonated
lima bean plants and ozonated lima bean plants simultaneously damaged by T.
urticae (Vuorinen et al.. 2004) however, other tritrophic interactions have shown no
effect (Pinto et al.. 2007b).
There are few studies that consider host location behaviors of parasites under
elevated O3. In closed chambers fumigated with O3, the searching efficiency and
proportion of the host larval fruit flies parasitized by Asobara tabida declined when
compared to filtered air controls (Gate et al., 1995). The host location behavior and
rate of parasitism of the wasp (Coesia plutellae) on Plutella xylostella-infested potted
cabbage plants was tested under ambient and doubled O3 conditions in an open-air
fumigation system (Pinto et al.. 2008). The number of wasps found in the field and
the percentages of parasitized larvae were not significantly different from controls
under elevated O3.
Elevated O3 has the potential to perturb specialized food-web communication in
transgenic crops. In insect-resistant oilseed rape Brassica napus grown under
100 ppb O3 in a growth chamber, reduced feeding damage by Putella xylostella led
to deceased attraction of the endoparasitoid (Costesia vestalis), however this
9-97
-------
tritrophic interaction was influenced by the degree of herbivore feeding (Himanen et
al., 2009a; Himanen et al., 2009b). Under chronic O3-exposure, the insect resistance
trait BT cry 1 Ac in transgenic B. napus was higher than the control (Himanen et al.,
2009c). There was a negative relative growth rate of the Bt target herbivore, P.
xylostella, in all O3 treatments.
9.4.9.4 Summary
Recent information on O3 effects on insects and other wildlife is limited to a few
species and there is no consensus on how these organisms respond to elevated O3.
Studies published since the last review show impacts of elevated O3 on both species-
level responses (reproduction, growth, feeding behavior) and community and
ecosystem-level responses (population growth, abundance, shift in community
structure) in some insects and soil fauna. Changes in ecologically important
behaviors such as feeding and thermoregulation have recently been observed with O3
exposure in amphibians and reptiles, however, these responses occur at
concentrations of O3 much higher than ambient levels.
Recent information available since the last review considers the effects of O3 on
chemical signaling in insect and wildlife interactions. Specifically, studies on O3
effects on pollination and seed dispersal, defenses against herbivory and predator-
prey interactions all consider the ability of O3 to alter the chemical signature of
VOCs emitted during these pheromone-mediated events. The effects of O3 on
chemical signaling between plants, herbivores and pollinators as well as interactions
between multiple trophic levels is an emerging area of study that may result in
further elucidation of O3 effects at the species, community and ecosystem-level.
9.5 Effects-based Air Quality Exposure Indices and Dose
Modeling
9.5.1 Introduction
Exposure indices are metrics that quantify exposure as it relates to measured plant
response (e.g., reduced growth). They are summary measures of monitored ambient
O3 concentrations over time, intended to provide a consistent metric for reviewing
and comparing exposure-response effects obtained from various studies. Such indices
may also provide a basis for developing a biologically-relevant air quality standard
for protecting vegetation and ecosystems. Effects on plant growth and/or yield have
been a major focus of the characterization of O3 impacts on plants for purposes of the
air quality standard setting process (U.S. EPA, 2007b, 1996e, 1986). The relationship
of O3 and plant responses can be characterized quantitatively as "dose-response" or
"exposure-response." The distinction is in how the pollutant concentration is
9-98
-------
expressed: "dose" is the pollutant concentration absorbed by the leaf over some time
period, and is very difficult to measure directly, whereas "exposure" is the ambient
air concentration measured near the plant over some time period, and summarized for
that period using an index. Exposure indices have been most useful in considering
the form of the secondary O3 NAAQS, in large part because they only require
ambient air quality data rather than more complex indirect calculations of dose to the
plant. The attributes of exposure indices that are most relevant to plant response are
the weighting of O3 concentrations and the daily and seasonal time-periods. Several
different types of exposure indices are discussed in Section 9.5.2.
From a theoretical perspective, a measure of plant O3 uptake or dose from ambient
air (either rate of uptake or cumulative seasonal uptake) might be a better predictor of
plant response to O3 than an exposure index and may be useful in improving risk
assessment. An uptake estimate would have to integrate all those environmental
factors that influence stomatal conductance, including but not limited to temperature,
humidity, and soil water status (Section 9.5.4). Therefore, uptake values are generally
obtained with simulation models that require knowledge of species- and site-specific
values for the variables mentioned. However, a limitation of modeling dose is that
environmental variables are poorly characterized. In addition, it has also been
recognized that O3 detoxification processes and the temporal dynamics of
detoxification must be taken into account in dose modeling (Heath et al., 2009)
(Section 9.5.4). Because of this, research has focused historically on predictors of O3
damage to plants based only on exposure as a summary measure of monitored
ambient pollutant concentration over some integral of time, rather than dose (U.S.
EPA. 1996c: Costa etal.. 1992: Lee et al.. 1988b: U.S. EPA. 1986: Lefohn and
Benedict. 1982: O'Gara. 1922).
9.5.2 Description of Exposure Indices Available in the Literature
Mathematical approaches for summarizing ambient air quality information in
biologically meaningful forms for O3 vegetation effects assessment purposes have
been explored for more than 80 years (U.S. EPA. 1996b: O'Gara. 1922). In the
context of national standards that protect for "known or anticipated" effects on many
plant species in a variety of habitats, exposure indices provide a numerical summary
of very large numbers of ambient observations of concentration over extended
periods. Like any summary statistic, exposure indices retain information on some,
but not all, characteristics of the original observations. Several indices have been
developed to attempt to incorporate some of the biological, environmental, and
exposure factors that influence the magnitude of the biological response and
contribute to observed variability (Hogsett et al., 1988). In the 1996 O3 AQCD (U.S.
EPA, 1996a), the exposure indices were arranged into five categories; (1) One event,
(2) Mean, (3) Cumulative, (4) Concentration weighted, and (5) Multicomponent, and
were discussed in detail (Lee et al., 1989). Figure 9-9 illustrates how several of the
indices weight concentration and accumulate exposure. For example, the SUM06
index (panel a) is a threshold-based approach wherein concentrations below
9-99
-------
0.06 ppm are given a weight of zero and concentrations at or above 0.06 ppm are
given a weight of 1.0 that is summed, usually over 3 to 6 months. The Sigmoid
approach (panel b), which is similar to the W126 index (Lefohn et al., 1988; Lefohn
and Runeckles, 1987), is a non-threshold approach wherein all concentrations are
given a weight that increases from zero to 1.0 with increasing concentration and
summed.
0.15
0.10
0.05
o.oo
c. 2HDM and M-7
.115
.070
ppm
2ndHDM-
M-7 =0.05 ppm
2 46 8 10 0
Day
246
Day
8 10
(a) SUM06: the upper graphic (within panel a) illustrates an episodic exposure profile; the shaded area under some of the peaks
illustrates the concentrations greater than or equal to 0.06 ppm that are accumulated in the index. The insert shows the
concentration weighting (0 or 1) function. The lower portion of panel a graphically illustrates how concentration is accumulated
over the exposure period, (b) SIGMOID: the upper graphic illustrates an episodic exposure profile; the variable shaded area under
the peaks illustrates the concentration-dependent weights that are accumulated in the index. The insert shows the sigmoid
concentration weighting function. This is similar to the W126 function. The lower portion of the graphic illustrates how
concentration is accumulated over the exposure period, (c) second HDM and M-7: the upper graphic illustrates an episodic
exposure profile. The lower portion of the graphic illustrates that the second HDM considers only a single exposure peak, while
the M-7 (average of 7-h daily means) applies a constant exposure value over the exposure period.
Source: Reprinted with permission of Air and Waste Management Association (Tingev et al.. 1991).
Figure 9-9 Diagrammatic representation of several exposure indices
illustrating how they weight concentration and accumulate
exposure.
9-100
-------
This section will primarily discuss SUM06, W126 and AOTx exposure metrics.
Below are the definitions of the three cumulative index forms:
• SUM06: Sum of all hourly O3 concentrations greater than or equal to
0.06 ppm observed during a specified daily and seasonal time window
(Figure 9-9. Panel A).
• AOTx: Sum of the differences between hourly O3 concentrations greater than
a specified threshold during a specified daily and seasonal time window. For
example, AOT40 is sum of the differences between hourly concentarions
above 0.04 ppm.
• W126: Sigmoidally weighted sum of all hourly O3 concentrations observed
during a specified daily and seasonal time window (Lefohn et al.. 1988;
Lefohn and Runeckles. 1987), similar to Figure 9-9, Panel B. The sigmoidal
weighting of hourly O3 concentration is given in the equation below, where C
is the hourly O3 concentration in ppm:
W =
c l + 4403e-126C
Equation 9-1
These indices have a variety of relevant time windows that may be applied and are
discussed in Section 9.5.3.
Various factors with known or suspected bearing on the exposure-response
relationship, including concentration, time of day, respite time, frequency of peak
occurrence, plant phenology, predisposition, etc., have been weighted with various
functions in a large set of indices. The resulting indices were evaluated by ranking
them according to the goodness-of-fit of a regression model of growth or yield
response (Lee et al.. 1989). The statistical evaluations for each of these indices were
completed using growth or yield response data from many earlier exposure studies
(e.g., NCLAN). This retrospective approach was necessary because there were no
studies specifically designed to test the goodness-of-fit of the various indices.
The goodness-of-fit of a set of linear and nonlinear models for exposure-response
was ranked as various proposed indices were used in turn to quantify exposure. This
approach provided evidence for the best indices. The results of retrospective analyses
are described below.
Most of the early retrospective studies reporting regression approaches used data
from the NCLAN program or data from Corvallis, Oregon or California (Costa et al..
1992: Leeetal. 1988b: Lefohn et al.. 1988: Musselman et al.. 1988: Leeetal. 1987:
U.S. EPA. 1986). These studies were previously reviewed by the EPA (U.S. EPA.
1996c: Costa et al.. 1992) and were in general agreement that the best fit to the data
resulted from using cumulative concentration-weighted exposure indices
(e.g., W126, SUM06). Lee et al. (1987) suggested that exposure indices that included
all the 24-h data performed better than those that used only 7 hours of data; this was
9-101
-------
consistent with the conclusions of Heagle et al. (1987) that plants receiving
exposures for an additional 5 hours/day showed 10% greater yield loss than those
exposed for 7 hours/day. In an analysis using the National Crop Loss Assessment
Network (NCLAN) data, Lee et al.(1988b) found several indices which only
cumulated and weighted higher concentrations (e.g., W126, SUM06, SUM08, and
AOT40) performed very well. Amongst this group no index had consistently better
fits than the other indices across all studies and species (Heagle et al.. 1994b: Lefohn
etal.. 1988: Musselman et al.. 1988). Lee et al. (1988b) found that adding phenology
weighting to the index somewhat improved the performance of the indices.
The "best" exposure index was a phenologically weighted cumulative index, with
sigmoid weighting on concentration and a gamma weighting function as a surrogate
for plant growth stage. This index provided the best statistical fit when used in the
models under consideration, but it required data on species and site conditions,
making specification of weighting functions difficult for general use.
Other factors, including predisposition time (Hogsett et al., 1988; McCool et al.,
1988) and crop development stage (Tingey et al., 2002; Heagle et al., 1991)
contributed to variation in the biological response and suggested the need for
weighting O3 concentrations to account for predisposition time and phenology.
However, the roles of predisposition and phenology in plant response vary
considerably with species and environmental conditions; therefore, specification of a
weighting function for general use in characterizing plant exposure has not been
possible.
European scientists took a similar approach in developing indices describing growth
and yield loss in crops and tree seedlings, using OTCs with modified ambient
exposures, but many fewer species and study locations were employed in the
European studies. There is evidence from some European studies that a lower (Pleijel
et al., 1997) or higher (Finnan et al., 1997; Finnan et al., 1996) cutoff value in indices
with a threshold may provide a better statistical fit to the experimental data. Finnan et
al. (1997) used seven exposure studies of spring wheat to confirm that cumulative
exposure indices emphasizing higher O3 concentrations were best related to plant
response and that cumulative exposure indices using weighting functions, including
cutoff concentrations, allometric and sigmoidal, provided a better fit and that the
ranking of these indices differed depending on the exposure-response model used.
Weighting those concentrations associated with sunshine hours in an attempt to
incorporate an element of plant uptake did not improve the index performance
(Finnan et al., 1997). A more recent study using data from several European studies
of Norway spruce, analyzed the relationship between relative biomass accumulation
and several cumulative, weighted indices, including the AOT40 (area over a
threshold of 40ppb) and the SUM06 (Skarby et al., 2004). All the indices performed
relatively well in regressing biomass and exposure index, with the AOT20 and
AOT30 doing slightly better than others (r2 = 0.46-0.47). In another comparative
study of four independent data sets of potato yield and different cumulative uptake
indices with different cutoff values, a similarly narrow range of r2 was observed
(r2 = 0.3-0.4) (Pleiiel et al.. 2004b).
9-102
-------
In Europe, the cutoff concentration-weighted index AOT40 was selected in
developing exposure-response relationships based on OTC studies of a limited
number of crops and trees (Grunhage and Jager, 2003). The United Nations
Economic Commission for Europe (TJNECE, 1988) adopted the critical levels
approach for assessment of O3 risk to vegetation across Europe. As used by the
UNECE, the critical levels are not like the air quality regulatory standards used in the
U.S., but rather function as planning targets for reductions in pollutant emissions to
protect ecological resources. Critical levels for O3 are intended to prevent long-term
deleterious effects on the most sensitive plant species under the most sensitive
environmental conditions, but not intended to quantify O3 effects. A critical level
was defined as "the concentration of pollutant in the atmosphere above which direct
adverse effects on receptors, such as plants, ecosystems, or materials may occur
according to present knowledge" (UNECE. 1988). The nature of the "adverse
effects" was not specified in the original definition, which provided for different
levels for different types of harmful effect (e.g., visible injury or loss of crop yield).
There are also different critical levels for crops, forests, and semi-natural vegetation.
The caveat, "according to present knowledge" is important because critical levels are
not rigid; they are revised periodically as new scientific information becomes
available. For example, the original critical level for O3 specified concentrations for
three averaging times, but further research and debate led to the current critical level
being stated as the cumulative exposure (concentration x hours) over a cutoff
concentration of 40 ppb (AOT40) (Fuhrer et al.. 1997).
More recently in Europe, a decision was made to work toward a flux-based approach
(see Section 9.5.4) for the critical levels ("Level II"), with the goal of modeling O3
flux-effect relationships for three vegetation types: crops, forests, and semi-natural
vegetation (Grunhage and Jager. 2003). Progress has been made in modeling flux
(U.S. EPA. 2006b) and the Mapping Manual is being revised (Ashmore et al.. 2004a.
b; Grennfelt 2004: Karlsson et al.. 2003). The revisions may include a flux-based
approach for three crops: wheat, potatoes, and cotton. However, because of a lack of
flux-response data, a cumulative, cutoff concentration-based (AOTx) exposure index
will remain in use for the near future for most crops and for forests and semi-natural
herbaceous vegetation (Ashmore et al.. 2004b).
In both the U.S. and Europe, the adequacy of these numerical summaries of exposure
in relating biomass and yield changes have, for the most part, all been evaluated
using data from studies not necessarily designed to compare one index to another
(Skarbv et al.. 2004: LeeetaL 1989: LefohnetaL 1988). Very few studies in the
U.S. have addressed this issue since the 2006 O3 AQCD. McLaughlin et al. (2007a)
reported that the cumulative exposure index of AOT60 related well to reductions in
growth rates at forest sites in the southern Appalachian Mountains. However, the
authors did not report an analysis to compare multiple indices. Overall, given the
available data from previous O3 AQCDs and the few recent studies, the cumulative,
concentration-weighted indices perform better than the peak or mean indices. It is
still not possible, however, to distinguish the differences in performance among the
cumulative, concentration-weighted indices.
9-103
-------
The main conclusions from the 1996 and 2006 O3 AQCDs regarding an index based
on ambient exposure are still valid. No information has come forth since the 2006 O3
AQCD to alter those conclusions. These key conclusions can be restated as follows:
• ozone effects in plants are cumulative;
• higher O3 concentrations appear to be more important than lower
concentrations in eliciting a response;
• plant sensitivity to O3 varies with time of day and plant development stage;
• quantifying exposure with indices that accumulate the O3 hourly
concentrations and preferentially weight the higher concentrations improves
the explanatory power of exposure/response models for growth and yield, over
using indices based on mean and peak exposure values.
Following the 2006 criteria review process (U.S. EPA, 2006b), the EPA proposed an
alternative form of the secondary NAAQS for O3 using a cumulative, concentration-
weighted exposure index to protect vegetation from damage (72 FR 37818).
The EPA considered two specific concentration-weighted indices: the cutoff
concentration weighted SUM06 and the sigmoid-weighted W126 exposure index
(U.S. EPA, 2007b). These two indices performed equally well in predicting the
exposure-response relationships observed in the crop and tree seedlings studies (Lee
et al, 1989). At a workshop convened to consider the science supporting these
indices (Heck and Cowling, 1997) there was a consensus that these cumulative
concentration-weighted indices being considered were equally capable of predicting
plant response. It should be noted that there are some important differences between
the SUM06 and W126. When considering the response of vegetation to O3 exposures
represented by the threshold (e.g., SUM06) and non-threshold (e.g., W126) indices,
the W126 metric does not have a cut-off in the weighting scheme as does SUM06
and thus it includes consideration of potentially damaging exposures below 60 ppb.
The W126 metric also adds increasing weight to hourly concentrations from about
40 ppb to about 100 ppb (Lefohn et al.. 1988: Lefohn and Runeckles. 1987). This is
unlike cut-off metrics such as the SUM06 where all concentrations above 60 ppb are
treated equally. This is an important feature of the W126 since as hourly
concentrations become higher, they become increasingly likely to overwhelm plant
defenses and are known to be more detrimental to vegetation (see Section 9.5.3.1).
9.5.3 Important Components of Exposure Indices
In the previous O3 AQCDs it was established that higher hourly concentrations have
greater effects on vegetation than lower concentrations (U.S. EPA, 2006b, 1996c).
Further, it was determined that the diurnal and seasonal duration of exposure is
important for plant response. Weighting of hourly concentrations and the diurnal and
seasonal time window of exposure are the most important variables in a cumulative
exposure index and will be discussed below. However, these variables should be
looked at in the context of plant phenology, diurnal conductance rates, plant canopy
structure, and detoxification mechanisms of vegetation as well as the climate and
9-104
-------
meteorology, all of which are determinants of plant response. These more specific
factors will be discussed in the uptake and dose modeling Section 9.5.4.
9.5.3.1 Role of Concentration
The significant role of peak O3 concentrations was established based on several
experimental studies (U.S. EPA, 1996c). Several studies (Oksanen and Holopainen,
2001; Yun and Laurence, 1999; Nussbaum et al., 1995) have added support for the
important role that peak concentrations, as well as the pattern of occurrence, plays in
plant response to O3. Oksanen and Holopainen (2001) found that the peak
concentrations and the shape of the O3 exposure (i.e., duration of the event) were
important determinants of foliar injury in European white birch saplings, but growth
reductions were found to be more related to total cumulative exposure. Based on air
quality data from 10 U.S. cities, three 4-week exposure treatments having the same
SUM06 value were constructed by Yun and Laurence (1999). The authors used
different exposure regimes to explore effects of treatments with variable versus
uniform peak occurrence during the exposure period. The authors reported that the
variable peak exposures were important in causing injury, and that the different
exposure treatments, although having the same SUM06, resulted in very different
patterns of foliar injury. Nussbaum et al. (1995) also found peak concentrations and
the pattern of occurrence to be critical in determining the measured response.
The authors recommended that to describe the effect on total forage yield, peak
concentrations >0.11 ppm must be emphasized by using an AOT with higher
threshold concentrations.
A greater role for peak concentrations in effects on plant growth might be inferred
based on air quality analyses for the southern California area (Tingey et al., 2004;
Lee et al., 2003a). In the late 1960s and 1970s, extremely high O3 concentrations had
impacted the San Bernardino National Forest. However, over the past 20+ years,
significant reductions in O3 exposure have occurred (Bytnerowicz et al., 2008; Lee et
al., 2003a; Lefohn and Shadwick, 2000; Davidson, 1993). An illustration of this
improvement in air quality is shown by the 37-year history of O3 air quality at the
Crestline site in the San Bernardino Mountains (Figure 9-10) (Lee et al., 2003a).
Ozone exposure increased from 1963 to 1979 concurrent with increased population
and vehicular miles, followed by a decline to the present mirroring decreases in
precursor emissions. The pattern in exposure was evident in various exposure indices
including the cumulative concentration weighted (SUM06), as well as maximum
peak event (1-h peak), and the number of days having hourly averaged O3
concentrations greater than or equal to 95 ppb. The number of days having hourly
averaged O3 concentrations greater than or equal to 95 ppb declined significantly
from 163 days in 1978 to 103 days in 1997. The changes in ambient O3 air quality
for the Crestline site were reflected in the changes in frequency and magnitude of the
peak hourly concentration and the duration of exposure (Figure 9-10). Considering
the role of exposure patterns in determining response, the seasonal and diurnal
9-105
-------
patterns in hourly O3 concentration did not vary appreciably from year to year over
the 37-year period (Lee et al., 2003a).
The potential importance of exposure to peak concentrations comes both from results
of measures of tree conditions on established plots and from results of model
simulations. Across a broad area of the San Bernardino National Forest, the Forest
Pest Management (FPM) method of injury assessment indicated an improvement in
crown condition from 1974 to 1988; and the area of improvement in injury
assessment is coincident with an improvement in O3 air quality (Miller and Rechel,
1999). A more recent analysis of forest changes in the San Bernardino National
Forest, using an expanded network of monitoring sites, has verified significant
changes in growth, mortality rates, basal area, and species composition throughout
the area since 1974 (Arbaugh et al., 2003). A model simulation of ponderosa pine
growth over the 40-year period in the San Bernardino National Forest showed a
significant impact of O3 exposure on tree growth and indicates improved growth
with reduced O3 concentrations. This area has also experienced elevated
N deposition and based on a number of environmental indicators, it appears that this
area is experiencing N saturation (Fenn et al., 1996). To account for this potential
interaction, the model simulations were conducted under conditions of unlimited soil
N. The actual interactions are not known. The improvement in growth over the years
was attributed to improved air quality, but no distinction was made regarding the
relative role of "mid-range" and higher hourly concentrations, only that improved
growth tracked decreasing SUM06, maximum peak concentration, and number of
days of hourly O3 >95 ppb (Tingey et al., 2004). A summary of air quality data from
1980 to 2000 for the San Bernardino National Forest area of the number of "mid-
range" hourly concentrations indicated no dramatic changes over this 20-year period,
ranging from about 1,500 to 2,000 hours per year (Figure 9-11). There was a slow
increase in the number of "mid-range" concentrations from 1980 to 1986, which
corresponds to the period after implementation of the air quality standard. Another
sharper increase was observed in the late 1990s. This pattern of occurrence of
mid-range hourly concentrations suggests a lesser role for these concentration ranges
compared to the higher values in either of the ground-level tree injury observations
of the model simulation of growth over the 40-year period.
9-106
-------
Łxo
o 8 in
sr
sis
T-
N
a-"
§8
1- I")
04
If
D a
W3S
""—0
r1
8
^ »
S2
^Ł0
Ł-0
,1 CM
O
- KI
ia -Q T-
xa
a a
"?§'
zH
o
• ROG , ••^.^
0 NOx • ,0__>-
-------
Crestline, San Bernardino, CA
Number of Hours 50 - 89 ppb
060710005
2
o
E
3
2500
2000
1500
1000
500
o
00
CM
00
O
00
tO
00
CO
CO
CM
cn
CD
0>
0
CO
o
o
o
CM
Year
Figure 9-11 The number of hourly average concentrations between 50 and
89 ppb for the period 1980-2000 for the Crestline, San Bernardino
County, CA, monitoring site.
9.5.3.2 Diurnal and Seasonal Exposure
Diurnal Exposure
The diurnal patterns of maximal leaf/needle conductance and occurrence of higher
ambient concentrations can help determine which hours during the day over a season
should be included in an exposure index. Stomatal conductance is species and
phenology dependent and is linked to both diurnal and seasonal meteorological
activity as well as to soil/site conditions (e.g., VPD, soil moisture). Daily patterns of
leaf/needle conductance are often highest in midmorning, whereas higher ambient O3
concentrations generally occur in early to late afternoon when stomata are often
partially closed and conductances are lower. Total O3 flux depends on atmospheric
and boundary layer resistances, both of which exhibit variability throughout the day.
Experimental studies with tree species demonstrated the decoupling of ambient O3
exposure, peak occurrence, and gas exchange, particularly in areas of drought
(Panek. 2004). Several studies have suggested that ponderosa pine trees in the
southern and northern Sierra Nevada Mountains may not be as susceptible to high O3
concentrations as to lower concentrations, due to reduced needle conductance and O3
uptake during the period when the highest concentrations occur (Panek et al.. 2002:
Panek and Goldstein. 2001: Bauer et al.. 2000: Arbaugh et al.. 1998). Panek et al.
9-108
-------
(2002) compared direct O3 flux measurements into a canopy of ponderosa pine and
demonstrated a lack of correlation of daily patterns of conductance and O3
occurrence, especially in the late season drought period; the authors concluded that a
consideration of climate or season was essential, especially considering the role of
soil moisture and conductance/uptake. In contrast, Grulke et al. (2002) reported high
conductance when O3 concentrations were high in the same species, but under
different growing site conditions. The longer-term biological responses reported by
Miller and Rechel (1999) for ponderosa pine in the same region, and the general
reduction in recent years in ambient O3 concentrations, suggest that stomatal
conductance alone may not be a sufficient indicator of potential vegetation injury or
damage. Another consideration for the effect of O3 uptake is the diurnal pattern of
detoxification capacity of the plant. The detoxification capacity may not follow the
same pattern as stomatal conductance (Heath et al.. 2009).
The use of a 12-h (8:00 a.m. to 8:00 p.m.) daylight period for a W126 cumulating
exposure was based primarily on evidence that the conditions for uptake of O3 into
the plant occur mainly during the daytime hours. In general, plants have the highest
stomatal conductance during the daytime and in many areas atmospheric turbulent
mixing is greatest during the day as well (Uddling et al., 2010; U.S. EPA, 2006b).
However, notable exceptions to maximum daytime conductance are cacti and other
plants with crassulacean acid metabolism (CAM photosynthesis) which only open
their stomata at night. This section will focus on plants with C3 and C4
photosynthesis, which generally have maximum stomatal conductance during the
daytime.
Recent reviews of the literature reported that a large number of species had varying
degrees of nocturnal stomatal conductance (Caird et al.. 2007: Dawson et al.. 2007:
Musselman and Minnick, 2000). The reason for night-time water loss through
stomata is not well understood and is an area of active research (e.g., Christman et
al., 2009; Howard et al., 2009). Night-time stomatal opening may be enhanced by O3
damage that could result in loss of stomatal control, and less complete closure of
stomata, than under low O3 conditions (Caird et al., 2007; Grulke et al., 2007b).
In general, the rate of stomatal conductance at night is much lower than during the
day (Caird et al., 2007). Atmospheric turbulence at night is also often low, which
results in stable boundary layers and unfavorable conditions for O3 uptake into
vegetation (Finkelstein et al., 2000). Nevertheless, nocturnal turbulence does
intermittently occur and may result in non-negligible O3 flux into the plants.
In addition, plants might be more susceptible to O3 exposure at night than during the
daytime, because of potentially lower plant defenses (Heath et al., 2009; Loreto and
Fares, 2007; Musselman et al., 2006; Musselman and Minnick, 2000). For significant
nocturnal stomatal flux and O3 effects to occur, specific conditions must exist.
A susceptible plant with nocturnal stomatal conductance and low defenses must be
growing in an area with relatively high night-time O3 concentrations and appreciable
nocturnal atmospheric turbulence. It is unclear how many areas there are in the U.S.
where these conditions occur. It may be possible that these conditions exist in
mountainous areas of southern California, front-range of Colorado (Turnipseed et al.,
2009) and the Great Smoky Mountains of North Carolina and Tennessee. Tobiessen
9-109
-------
(1982) found that shade intolerant tree species showed opening of stomata in the dark
and did not find this in shade tolerant species. This may indicate shade intolerant
trees may be more likely to be susceptible to O3 exposure at night. More information
is needed in locations with high night-time O3 to assess the local O3 patterns,
micrometeorology and responses of potentially vulnerable plant species.
Several field studies have attempted to quantify night-time O3 uptake with a variety
of methods. However, many of these studies have not linked the night-time flux to
measured effects on plants. Grulke et al. (2004) showed that the stomatal
conductance at night for ponderosa pine in the San Bernardino National Forest (CA)
ranged from one tenth to one fourth that of maximum daytime stomatal conductance.
In June, at a high-elevation site, it was calculated that 11% of the total daily O3
uptake of pole-sized trees occurred at night. In late summer, however, O3 uptake at
night was negligible. However, this study did not consider the turbulent conditions at
night. Finkelstein et al. (2000) investigated O3 deposition velocity to forest canopies
at three different sites. The authors found the total flux (stomatal and non-stomatal)
to the canopy to be very low during night-time hours as compared to day-time hours.
However, the authors did note that higher nocturnal deposition velocities at conifer
sites may be due to some degree of stomatal opening at night (Finkelstein et al.,
2000). Work by Mereu et al. (2009) in Italy on Mediterranean species indicated that
nocturnal uptake was from 10 to 18% of total daily uptake during a weak drought
and up to 24% as the drought became more pronounced. The proportion of night-
time uptake was greater during the drought due to decreases in daytime stomatal
conductance (Mereu et al., 2009). In a study conducted in California, (Fares et al.,
2011) reported that calculated mean percentages of nocturnal uptake were 5%,
12.5%, 6.9% of total O3 uptake for lemon, mandarin, and orange, respectively.
In another recent study at the Aspen FACE site in Wisconsin, calculated leaf-level
stomatal O3 flux was near zero from the night-time hours of 8:00 p.m. to 5:00 a.m.
(Uddling et al.. 2010). This was likely due to low horizontal wind speed (>1
meter/sec) and low O3 concentrations (<25 ppb) during those same night-time hours
(Figure 9-12).
9-110
-------
E
j-:
O
"o
0.15
o.io
0.05
0.00
2004
2005 Sf
10 15
Time of day
SO
Ł 40
"o
I 30
.1
1 20
s
cf
5 10 15
Time of day
u.
•a 2
to
10 15
Time of day
T 6
W
•51
E 5
o
I «
•0
CO
ID 15
Time of day
Note: Subscripts "max" and "min" refer to stomatal fluxes calculated neglecting and accounting for potential non-stomatal O3 flux,
respectively.
Source: Reprinted with permission of Elsevier Ltd. (Uddling et al.. 2010).
Figure 9-12 Diurnal (a) conductance through boundary layer and stomata (gbs),
(b) ozone concentration, and leaf-level stomatal Os flux (FstOI) in
control plots from mid-June through August, in (c) 2004 and
(d) 2005 in the Aspen FACE experiment.
A few studies have tested the biological effects of night-time O3 exposure on
vegetation in controlled chambers. Biomass of ponderosa pine seedlings was
significantly reduced when seedlings were exposed to either daytime or nighttime
episodic profiles (Lee and Hogsett 1999). However, the biomass reductions were
much greater with daytime peak concentrations than with nighttime peak
concentrations. Similarly, birch cuttings grown in field chambers that were exposed
to O3 at night only, daytime only, and 24 hours showed similar reductions in biomass
in night only and day only treatments. Birch seedling showed greater reductions in
growth in 24-h exposures than those exposed to O3 at night or day only (Matvssek et
al.. 1995). Field mustard (Brassica rapd) plants exposed to O3 during the day or
night showed little significant difference in the amounts of injury or reduced growth
response to O3 treatment, although the stomatal conductance was 70-80% lower at
9-111
-------
night (Winner et al., 1989). These studies show that effects can be seen with night-
time exposures to O3 but when atmospheric conditions are stable at night, it is
uncertain how these exposures may affect plants and trees with complex canopies in
the field.
Seasonal exposure
Vegetation across the U.S. has widely varying periods of physiological activity
during the year due to variability in climate and phenology. In order for a particular
plant to be vulnerable to O3 pollution, it must have foliage and be physiologically
active. Annual crops are typically grown for periods of two to three months.
In contrast, perennial species may be photosynthetically active longer (up to
12 months each year for some species) depending on the species and where it is
grown. In general, the period of maximum physiological activity and thus, potential
O3 uptake for vegetation coincides with some or all of the intra-annual period
defined as the O3 season, which varies on a state-by-state basis (Figure 3-24). This is
because the high temperature and high light conditions that typically promote the
formation of tropospheric O3 also promote physiological activity in vegetation. There
are very limited exceptions to this pattern where O3 can form in the winter in areas in
the western U.S. with intense natural gas exploration (Pinto, 2009), but this is
typically when plants are dormant and there is little chance of O3 uptake. Given the
significant variability in growth patterns and lengths of growing season among the
wide range of vegetation species that may experience adverse effects associated with
O3 exposure, no single time window of exposure can work perfectly for all types of
vegetation.
Various intra-annual averaging and accumulation time periods have been considered
for the protection of vegetation. The 2007 proposal for the secondary O3 standard (75
FR 37818) proposed to use the maximum consecutive 3-month period within the O3
season. The U.S. Forest Service and federal land managers have used a 24-h W126
accumulated for 6 months from April through September (U.S. Forest Service,
2000). However, some monitors in the U.S. are operational for as little as four
months and would not have enough data for a 6-month seasonal window.
The exposure period in the vast majority of O3 exposure studies conducted in the
U.S. has been much shorter than 6 months. Most of the crop studies done through
NCLAN had exposures less than three months with an average of 77 days. Open-top
chamber studies of tree seedlings, compiled by the EPA, had an average exposure of
just over three months or 99 days. In more recent FACE experiments, SoyFACE
exposed soybeans for an average of approximately 120 days per year and the Aspen
FACE experiment exposed trees to an average of approximately 145 days per year of
elevated O3, which included the entire growing season at those particular sites.
Despite the possibility that plants may be exposed to ambient O3 longer than
3 months in some locations, there is generally a lack of exposure experiments
conducted for longer than 3 months.
9-112
-------
30 40
Highest 3 month W126
20
30 40 50
Highest 3 month W126
70
Note: Data are from the AQS and CASTNET monitors for the years 2008 and 2009. (A) W126, 3 month versus 6 month, 2008
(Pearson correlation = 0.99); (B) W126, 3 month versus 6 month, 2009 (Pearson correlation = 0.99).
Figure 9-13 Maximum 3-month, 12-h W126 plotted against maximum 6-month,
12-hW126.
In an analysis of the 3- and 6-month maximum W126 values calculated for over
1,200 AQS (Air Quality System) and CASTNET (Clean Air Status and Trend
Network) EPA monitoring sites for the years 2008-2009, it was found that these 2
accumulation periods resulted in highly correlated metrics (Figure 9-13). The two
accumulation periods were centered on the yearly maximum for each monitoring site,
and it is possible that this correlation would be weaker if the two periods were not
temporally aligned. In the U.S., W126 cumulated over 3 months, and W126
9-113
-------
cumulated over 6 months are proxies of one another, as long as the period in which
daily W126 is accumulated corresponds to the seasonal maximum. Therefore, it is
expected that either statistic will predict vegetation response equally well. In other
words, the strength of the correlation between maximum 3-month W126 and
maximum 6-month W126 is such that there is no material difference in their
predictive value for vegetation response.
9.5.4 Ozone Uptake/Dose Modeling for Vegetation
Another approach for improving risk assessment of vegetation response to ambient
O3 is based on estimating the O3 concentration from the atmosphere that enters the
leaf (i.e., flux or deposition). Interest has been increasing in recent years, particularly
in Europe, in using mathematically tractable flux models for O3 assessments at the
regional, national, and European scale (Matyssek et al., 2008; Paoletti and Manning,
2007: ICP M&M. 2004: Emberson et al.. 2000b: Emberson et al.. 2000a). Some
researchers have claimed that using flux models can be used to better predict
vegetation responses to O3 than exposure-based approaches (Matyssek et al., 2008).
However, other research has suggested that flux models do not predict vegetation
responses to O3 better than exposure-based models, such as AOT40 (Gonzalez-
Fernandez et al., 2010). While some efforts have been made in the U.S. to calculate
O3 flux into leaves and canopies (Fares et al., 2010a: Turnipseed et al., 2009:
Uddling et al.. 2009: Bergweiler et al.. 2008: Hogg et al.. 2007: Grulke et al.. 2004:
Grantz et al., 1997: Grantz et al., 1995), little information has been published relating
these fluxes to effects on vegetation. The lack of flux data in the U.S. and the lack of
understanding of detoxification processes have made this technique less viable for
vulnerability and risk assessments in the U.S.
Flux calculations are data intensive and must be carefully implemented. Reducing
uncertainties in flux estimates for areas with diverse surface or terrain conditions to
within ± 50% requires "very careful application of dry deposition models, some
model development, and support by experimental observations" (Wesely and Hicks,
2000). As an example, the annual average deposition velocity of O3 among three
nearby sites in similar vegetation was found to vary by ± 10%, presumably due to
terrain (Brook et al., 1997). Moreover, the authors stated that the actual variation was
even greater, because stomatal uptake was unrealistically assumed to be the same
among all sites, and flux is strongly influenced by stomatal conductance (Brook et
al., 1997: Massman and Grantz, 1995: Fuentes et al., 1992: Reich, 1987: Leuning et
al., 1979). This uptake-based approach to quantify the vegetation impact of O3
requires inclusion of those factors that control the diurnal and seasonal O3 flux to
vegetation (e.g., climate patterns, species and/or vegetation-type factors and site-
specific factors). The models have to distinguish between stomatal and non-stomatal
components of O3 deposition to adequately estimate actual concentration reaching
the target tissue of a plant to elicit a response (Uddling et al., 2009). Determining this
O3 uptake via canopy and stomatal conductance relies on models to predict flux and
ultimately the "effective" flux (Grunhage et al., 2004: Massman, 2004: Massman et
9-114
-------
al.. 2000). "Effective flux" has been defined as the balance between O3 flux and
detoxification processes (Heath et al., 2009; Musselman and Massman. 1999;
Grunhage and Haenel. 1997; Dammgen et al., 1993). The time-integrated "effective
flux" is termed "effective dose." The uptake mechanisms and the resistances in this
process, including stomatal conductance and biochemical defense mechanisms, are
discussed below. The flux-based index is the goal for the "Level II" critical level for
assessment of O3 risk to vegetation and ecosystems across Europe (Ashmore et al..
2004a).
An important consideration in both O3 exposure and uptake is how the O3
concentration at the top of low vegetation such as, crops and tree seedlings may be
lower than the height at which the measurement is taken. Ambient monitor inlets in
the U.S. are typically at heights of 3 to 5 meters. During daytime hours, the vertical
O3 gradient can be relatively small because turbulent mixing maintains the
downward flux of O3. For example, Horvath et al. (1995) calculated a 7% decrease in
O3 going from a height of 4 meters down to 0.5 meters above the surface during
unstable (or turbulent) conditions in a study over low vegetation in Hungary [see
Section AX3.3.2. of the 2006 O3 AQCD (U.S. EPA. 2006b)1. There have been
several studies indicating decreased O3 concentrations under tree canopies (Kolb et
al.. 1997; Samuelson and Kelly. 1997; Joss and Graber. 1996; Fredericksen et al..
1995; Lorenzini andNali. 1995; Enders. 1992; Fontanetal.. 1992; Neufeld et al..
1992). In contrast, for forests, measured data may underestimate O3 concentration at
the top of the canopy. The difference between measurement height and canopy height
is a function of several factors, the intensity of turbulent mixing in the surface layer
and other meteorological factors, canopy height and total deposition to the canopy.
Some researchers have used deposition models to estimate O3 concentration at
canopy-top height based on concentrations at measurement height (Emberson et al..
2000a). However, deposition models usually require meteorological data inputs that
are not always available or well characterized across large geographical scales.
Soil moisture is a critical factor in controlling O3 uptake through its effect on plant
water status and stomatal conductance. In an attempt to relate uptake, soil moisture,
and ambient air quality to identify areas of potential risk, available O3 monitoring
data for 1983 to 1990 were used along with literature-based seedling exposure-
response data from regions within the southern Appalachian Mountains that might
have experienced O3 exposures sufficient to inhibit growth (Lefohn et al.. 1997). In a
small number of areas within the region, O3 exposures and soil moisture availability
were sufficient to possibly cause growth reductions in some O3 sensitive species
(e.g., black cherry). The conclusions were limited, however, because of the
uncertainty in interpolating O3 exposures in many of the areas and because the
hydrologic index used might not reflect actual water stress.
The non-stomatal component of plant defenses are the most difficult to quantify, but
some studies are available (Heath et al.. 2009; Barnes et al.. 2002; Plochl et al.. 2000;
Chen et al.. 1998; Massman and Grantz. 1995). Massman et al. (2000) developed a
conceptual model of a dose-based index to determine how plant injury response to
O3 relates to the traditional exposure-based parameters. The index used time-
9-115
-------
varying-weighted fluxes to account for the fact that flux was not necessarily
correlated with plant injury or damage. The model applied only to plant foliar injury
and suggested that application of flux-based models for determining plant damage
(yield or biomass) would require a better understanding and quantification of the
relationship between injury and damage.
9.5.5 Summary
Exposure indices are metrics that quantify exposure as it relates to measured plant
damage (i.e., reduced growth). They are summary measures of monitored ambient O3
concentrations over time intended to provide a consistent metric for reviewing and
comparing exposure-response effects obtained from various studies. No recent
information is available since 2006 that alters the basic conclusions put forth in the
2006 and 1996 O3 AQCDs. These AQCDs focused on the research used to develop
various exposure indices to help quantify effects on growth and yield in crops,
perennials, and trees (primarily seedlings). The performance of indices was
compared through regression analyses of earlier studies designed to support the
estimation of predictive O3 exposure-response models for growth and/or yield of
crops and tree (seedling) species.
Another approach for improving risk assessment of vegetation response to ambient
O3 is based on determining the O3 concentration from the atmosphere that enters the
leaf (i.e., flux or deposition). Interest has been increasing in recent years, particularly
in Europe, in using mathematically tractable flux models for O3 assessments at the
regional, national, and European scale (Matyssek et al.. 2008: Paoletti and Manning.
2007: TCP M&M. 2004: Emberson et al.. 2000b: Emberson et al.. 2000a). While
some efforts have been made in the U.S. to calculate O3 flux into leaves and canopies
(Turnipseed et al.. 2009: Uddling et al.. 2009: Bergweiler et al.. 2008: Hogg et al..
2007: Grulke et al.. 2004: Grantzetal.. 1997: Grantzetal.. 1995). little information
has been published relating these fluxes to effects on vegetation. There is also
concern that not all O3 stomatal uptake results in a yield reduction, which depends to
some degree on the amount of internal detoxification occurring with each particular
species. Those species having high amounts of detoxification potential may, in fact,
show little relationship between O3 stomatal uptake and plant response (Musselman
and Massman. 1999). The lack of data in the U.S. and the lack of understanding of
detoxification processes have made this technique less viable for vulnerability and
risk assessments in the U.S.
9-116
-------
The main conclusions from the 1996 and 2006 O3 AQCDs regarding indices based
on ambient exposure are still valid. These key conclusions can be restated as follows:
• Ozone effects in plants are cumulative;
• higher O3 concentrations appear to be more important than lower
concentrations in eliciting a response;
• plant sensitivity to O3 varies with time of day and plant development stage;
• quantifying exposure with indices that accumulate the O3 hourly
concentrations and preferentially weight the higher concentrations improves
the explanatory power of exposure/response models for growth and yield, over
using indices based on mean and peak exposure values.
Various weighting functions have been used, including threshold-weighted
(e.g., SUM06) and continuous sigmoid-weighted (e.g., W126) functions. Based on
statistical goodness-of-fit tests, these cumulative, concentration-weighted indices
could not be differentiated from one another using data from previous exposure
studies. Additional statistical forms for O3 exposure indices have been discussed in
Lee et al. (1988b). The majority of studies published since the 2006 O3 AQCD do
not change earlier conclusions, including the importance of peak concentrations, and
the duration and occurrence of O3 exposures in altering plant growth and yield.
Given the current state of knowledge and the best available data, exposure indices
that cumulate and differentially weight the higher hourly average concentrations and
also include the "mid-level" values continue to offer the most defensible approach
for use in developing response functions and comparing studies, as well as for
defining future indices for vegetation protection.
9.6 Ozone Exposure-Plant Response Relationships
9.6.1 Introduction
The adequate characterization of the effects of O3 on plants for the purpose of setting
air quality standards is contingent not only on the choice of the index used
(i.e., SUM06, W126) to summarize O3 concentrations (Section 9.5), but also on
quantifying the response of the plant variables of interest at specific values of the
selected index. The many factors that determine the response of plants to O3
exposure have been discussed in previous sections. They include species, genotype
and other genetic characteristics (Section 9.3), biochemical and physiological status
(Section 9.3), previous and current exposure to other stressors (Section 9.4.8), and
characteristics of the exposure itself (Section 9.5). Establishing a secondary air
quality standard entails the capability to generalize those observations, in order to
obtain predictions that are reliable enough under a broad variety of conditions, taking
into account these factors. This section reviews results that have related specific
quantitative observations of O3 exposure with quantitative observations of plant
9-117
-------
responses, and the predictions of responses that have been derived from those
observations through empirical models.
For four decades, exposure to O3 at ambient concentrations found in many areas of
the U.S. has been known to cause detrimental effects in plants (U.S. EPA. 2006K
1996b. 1984. 1978a). Results published after the 2006 O3 AQCD continue to support
this finding, and the following sections deal with the quantitative characterizations of
the relationship, and what new insights may have appeared since 2006. Detrimental
effects on plants include visible injury, decreases in the rate of photosynthesis,
reduced growth, and reduced yield of marketable plant parts. Most published
exposure-response data have reported O3 effects on the yield of crops and the growth
of tree seedlings, and those two variables have been the focus of the characterization
of ecological impacts of O3 for the purpose of setting secondary air quality standards.
In order to support quantitative modeling of exposure-response relationships, data
should preferably include more than three levels of exposure, and some control of
potential confounding or interacting factors should be present in order to model the
relationship with sufficient accuracy. Letting potential confounders, such as other
stressors, vary freely when generating O3 exposure-response data might improve the
'realism' of the data, but it also greatly increases the amount of data necessary to
extract a clear quantitative description of the relationship. Conversely however,
experimental settings should not be so exhaustively restrictive as to make
generalization outside of them problematic. During the last four decades, many of the
studies of the effects of O3 on growth and yield of plants have not included enough
levels of O3 to parameterize more than the simplest linear model. The majority of
these studies have only contrasted two levels, ambient and elevated, or sometimes
three by adding carbon filtration in OTC studies, with little or no consideration of
quantitatively relating specific values of exposure to specific values of growth or
yield. This is not to say that studies that did not include more than two or three levels
of O3 exposure, or studies that were conducted in uncontrolled environments, do not
provide exposure-response information that is highly relevant to reviewing air quality
standards. In fact, they can be essential in verifying the agreement between
predictions obtained through the empirical models derived from experiments such as
NCLAN, and observations. The consensus of model predictions and observations
from a variety of studies conducted in other locations, at other times, and using
different exposure methods, greatly increases confidence in the reliability of both.
Furthermore, if they are considered in the aggregate, studies with few levels of
exposure or high unaccounted variability can provide additional independent
estimates of decrements in plant growth and yield, at least within a few broad
categories of exposure.
Extensive exposure-response information on a wide variety of plant species has been
produced by two long-term projects that were designed with the explicit aim of
obtaining quantitative characterizations of the response of such an assortment of crop
plants and tree seedlings to O3 under North American conditions: the NCLAN
project for crops, and the EPA National Health and Environmental Effects Research
Laboratory, Western Ecology Division tree seedling project (NHEERL/WED).
The NCLAN project was initiated by the EPA in 1980 primarily to improve
9-118
-------
estimates of yield loss under field conditions and to estimate the magnitude of crop
losses caused by O3 throughout the U.S. (Heck et al.. 1991; Heck et al.. 1982).
The cultural conditions used in the NCLAN studies approximated typical agronomic
practices, and the primary objectives were: (1) to define relationships between yields
of major agricultural crops and O3 exposure as required to provide data necessary for
economic assessments and development of O3 NAAQS; (2) to assess the national
economic consequences resulting from O3 exposure of major agricultural crops; and
(3) to advance understanding of cause-and-effect relationships that determine crop
responses to pollutant exposures.
NCLAN experiments yielded 54 exposure-response curves for 12 crop species, some
of which were represented by multiple cultivars at several of 6 locations throughout
the United States. The NHEERL/WED project was initiated by EPA in 1988 with
similar objectives for tree species, and yielded 49 exposure-responses curves for
multiple genotypes of 11 tree species grown for up to three years in Oregon,
Michigan, and the Great Smoky Mountains National Park. Both projects used OTCs
to expose plants to three to five levels of O3. Eight of the 54 crop datasets were from
plants grown under a combination of O3 exposure and experimental drought
conditions. Figure 9-14 through Figure 9-17 summarizes some of the NCLAN and
NHEERL/WED results.
It should be noted that data from FACE experiments might also be used for modeling
exposure-response. They only use two levels of O3 (ambient concentration at the site
and a multiple of it), but given that the value of both levels of exposure changes
every year, and that they are typically run for many consecutive years, aggregating
data over time produces twice as many levels of O3 as there are years. As described
in Section 9.2.4. FACE experiments seek to impose fewer constraints on the growth
environment than OTCs. As a consequence, FACE studies have to contend with
larger variability, especially year-to-year variability, but the difference in
experimental conditions between the two methodologies makes comparisons between
their results especially useful.
Growth and yield of at least one crop (soybean) has been investigated in yearly
experiments since 2001 at a FACE facility in Illinois (UTUC. 2010; Morgan et al..
2006). However, almost all analyses of SoyFACE published so far have been based
on subsets of one or two years, and have only contrasted ambient versus elevated O3
as categorical variables. They have not modeled the response of growth and yield to
O3 exposure continuously over the range of exposure values that have occurred over
time. The only exception is a study by Betzelberger et al. (2010). who used a linear
regression model on data pooled over 2 years. Likewise, trees of three species
(trembling aspen, paper birch, and sugar maple) were grown between 1998 and 2009
in a FACE experiment located in Rhinelander, Wisconsin (Pregitzer et al.. 2008;
Dickson et al.. 2000). The Aspen FACE experiment has provided extensive data on
responses of trees beyond the seedling stage under long-term exposure, and also on
ecosystem-level responses (Section 9.4). but the only attempt to use those data in a
continuous model of the response of tree growth to O3 exposure (Percy et al.. 2007)
suffered from severe methodological problems, some of which are discussed in
9-119
-------
Section 9.6.3. Finally, one experiment was able to exploit a naturally occurring
gradient of O3 concentrations to fit a linear regression model to the growth of
cottonwood (Gregg et al.. 2006, 2003). Factors such as genotype, soil type and soil
moisture were under experimental control, and the authors were able to partition out
the effects of potential confounders such as temperature, atmospheric N deposition,
and ambient CO2.
A serious difficulty in assessing results of exposure-response research is the
multiplicity of O3 metrics that have been used in reporting. As described in Section
9.5, metrics that entail either weighting or thresholding of hourly values cannot be
algebraically converted into one another, or into unweighted metrics such as hourly
average. When computing O3 exposure using weighted or thresholded metrics, each
metric has to be computed separately from the original hourly data. Comparisons of
exposure-response models can only be made between studies that used the same
metric, and the value of exposure at which a given plant response is expected using
one metric of exposure cannot be exactly converted to another metric. Determining
the exposure value at which an effect would be observed in a different metric can
only be accomplished by first computing the experimental exposures in this metric
from the hourly data, then estimating (fitting) model coefficients again. This problem
is irremediable, although useful comparisons might be made using categorical
exposures such as 'current ambient exposure' or '2050 projected exposure', which
can serve as a common reference for quantitative values expressed in various metrics.
Studies that contained growth or yield exposure-response data at few levels of
exposure, and/or using metrics other than W126 are summarized in Table 9-17 and
Table 9-18.
9.6.2 Estimates of Crop Yield Loss and Tree Seedling Biomass Loss in the
1996 and 2006 Ozone AQCDs
The 1996 and 2006 O3 AQCDs relied extensively on analyses of NCLAN and
NHEERL/WED by Lee et al. (1994: 1989. 1988b. 1987). Hogsett et al. (1997). Lee
and Hogsett (1999). Heck et al. (1984). Rawlings and Cure (1985). Lesser et al.
(1990). and Gumpertz and Rawlings (1992). Those analyses concluded that a three-
parameter Weibull model -
Y= ae v >
Equation 9-2
is the most appropriate model for the response of absolute yield and growth to O3
exposure, because of the interpretability of its parameters, its flexibility (given the
small number of parameters), and its tractability for estimation. In addition, removing
the intercept a results in a model of relative yield (yield relative to [yield at
exposure=0]) without any further reparameterization. Formulating the model in terms
9-120
-------
of relative yield or relative yield loss (yield loss=[l - relative yield]) is essential in
comparing exposure-response across species, genotypes, or experiments for which
absolute values of the response may vary greatly. In the 1996 and 2006 O3 AQCDs,
the two-parameter model of relative yield was used in deriving common models for
multiple species, multiple genotypes within species, and multiple locations.
Given the disparate species, genotypes, and locations that were included in the
NCLAN and NHEERL/WED projects, and in the absence of plausible distributional
assumptions with respect to those variables, a three step process using robust
methods was used to obtain parameter estimates that could be generalized.
The models that were derived for each species or group of species were referred to as
median composite functions. In the first step, the three parameters of the Weibull
model were computed for absolute yield or biomass data from each NCLAN and
NHEERL/WED experiment (54 crop datasets and 49 tree seedling datasets), using
nonlinear regression. When data were only available for three levels of exposure
because of experimental problems, the shape parameter (3 was constrained to 1,
reducing the model to an exponential decay model. In the second step, a was
dropped, and predicted values of relative yield or biomass were then computed for
12-hour W126 exposures between 0 and 60 ppm-h. At each of these W126 exposure
values, the 25th, 50th, and 75th percentiles of the response were identified among the
predicted curves of relative response. For example, for the 34 NCLAN studies of 12
crop species grown under non-droughted conditions for a complete cropping cycle
(Figure 9-14), the 3 quartiles of the response were identified at every integer value of
W126 between 0 and 60. The third step fitted a two-parameter Weibull model to
those percentiles, yielding the median composite function for the relative yield or
biomass response to O3 exposure for each grouping of interest (e.g., all crops, all
trees, all datasets for one species), as well as composite functions for the other
quartiles. In the 1996 and 2006 O3 AQCDs this modeling of crop yield loss and tree
seedling biomass loss was conducted using the SUM06 metric for exposure. This
section updates those results by using the 12-hour W126 as proposed in 2007 (72 FR
37818) and 2010 (75 FR2938, page 3.003). Figure 9-14 through Figure 9-17 present
quantiles of predicted relative yield or biomass loss at seven values of the 12-h W126
for some representative groupings of NCLAN and NHEERL/WED results. Table 9-9
through Table 9-11 give the 90-day 12-h W126 O3 exposure values at which 10 and
20% yield or biomass losses are predicted in 50 and 75% of crop or tree species
using the composite functions.
9-121
-------
100 •
90 •
80 -
g 70 •
CO
o 60 -
_i
H 50 "
1 40 -
B
S. 30 -
20 •
10 -
0 -
c
34 crop datasets
T
^\^\
) 10 20 30 40
^
-r 90mPctile
i-L| 75'"Pctile
-
^
^~
•
50lhPctile
25'"Pctile
10lhPctile
50 60
12hrW126 (ccm-hr)
Note: Quantiles of the predicted relative yield loss at 7 values of 12-hour W126 for 34 Weibull curves estimated using nonlinear
regression on data from 34 studies of 12 crop species grown under well-watered conditions for the full duration of 1 cropping
cycle.
Source of Weibull parameters: Lee and Hogsett (1996).
Figure 9-14 Quantiles of predicted relative yield loss for 34 NCLAN crop
experiments.
9-122
-------
100 -
90 -
80 -
70 -
60 -
50 -
4° "
so -
20 -
10 -
0 -
11 Soybean datasets
i-L, 73»Pctile
10 20 30 40
100 -
90 -
1 50 -
§ 4° "
I 30 -
20 -
10 -
0
5 Cotton datasets
20 30 40 50
12hrW126 (ppm-hr
100 -
90 -
50 -
40 -
30 -
20 -
10 -
0
2 Com datasets
20 30 40 50
12hrW126 (ppm-hr)
Notes: Quantiles of the predicted relative yield loss at 7 values of 12-h W126 for Weibull curves estimated using nonlinear
regression for 4 species grown under well-watered conditions for the full duration of 1 cropping cycle. The number of studies
available for each species is indicated on each plot.
Source of Weibull parameters: Lee and Hogsett (1996).
Figure 9-15 Quantiles of predicted relative yield loss for 4 crop species in
NCLAN experiments.
9-123
-------
in
in
o
_i
in
in
ra
o
In
-------
100
90
80
I ™
I 6°
1 50
o
to 40
i
te 30
Q-
20
10
0
14 Aspendatasets
^- SCPPctile
100 -
90 -
80 -
70 -
60 -
50 -
40 -
30 -
20 -
10 -
0
11 Ponderosapine datasets
10 20 30 40 50 60
10 20 30 40 50 60
100 -
90 -
80 -
E 50 -
O
iS 40 -
I 30 -
Q_
20 -
10 -
0
7 Douglas firdatasets
20 30 40 50
90 day 12 hr W126 (ppm-hr)
100 -
90 -
80 -
70 -
60 -
50 -
40 -
30 -
20 -
10 -
STulip poplardatasets
10 20 30 40 50
90 day 12 hr W126 (ppm-hr)
Note: Quantiles of the predicted relative above-ground biomass loss at 7 exposure values of 12-h W126 for Weibull curves
estimated using nonlinear regression on data for 4 tree species grown under well-watered conditions for 1 or 2 year. Curves were
standardized to 90-day W126. The number of studies available for each species is indicated on each plot.
Source of Weibull parameters: Lee and Hogsett (1996).
Figure 9-17 Quantiles of predicted relative biomass loss for 4 tree species in
NHEERL/WED experiments.
9-125
-------
Table 9-9 Ozone exposures at which 10 and 20% yield loss is predicted for 50
and 75% of crop species.
Predited Yield Loss for 90-day 12-h W126 90-day 12-h W126
Crop Species^ for 10% yield loss (ppm-h) for 20% yield loss (ppm-h)
Model for the 50th Percentile of 34
curves
Relative 22 37
yield=exp(-(W126/104.82)**1.424)
Model for the 75th Percentile of 34
curves
Relative 16 27
yield=exp(-(W126/78.12)**1.415)
"Based on composite functions for the 50th and 75th percentiles of 34 Weibull curves for relative yield loss data from 34 non-
droughted NCLAN studies of 12 crop species; curves were standardized to 90-day W126.
Source of parameters for the 34 curves: Lee and Hogsett (1996).
Table 9-10 Ozone exposures at which 10 and 20% yield loss is predicted for 50
and 75% of crop species (Droughted versus Watered conditions).
90day12-hW126 90 day 12-h W126
Predicted Yield Loss for Crop Species3 for 10% yield loss (ppm-h) for 20% yield loss (ppm-h)
Model for the 50th Percentile of 2x8 curves
Watered Relative yield=exp(-(W126/132.86)**1.170) 19 37
Droughted Relative yield=exp(-(W126/179.84)**! .713) 48 75
Model for the 75th Percentile of 2><8 curves
Watered Relative yield=exp(-(W126/90.43)**! .310) 16 29
Droughted Relative yield=exp(-(W126/105.16)**! .833) 31 46
aUnder drought conditions and adequate moisture based on composite functions for the 50th and 75th percentiles of 16 Weibull
curves for relative yield loss data from 8 NCLAN studies that paired draughted and watered conditions for the same genotype;
curves were standardized to 90-day W126.
Source of parameters for the 16 curves: Lee and Hogsett (1996).
9-126
-------
Table 9-11 Ozone exposures at which 10 and 20% biomass loss is predicted
for 50 and 75% of tree species.
90day12hW126 90 day 12 h W126
Predicted Biomass Loss for Tree Species3 for 10% yield loss (ppm-h) for 20% yield loss (ppm-h)
Model for the 50th Percentile of 49 curves
Relative yield=exp(-(W126/131.57)**1.242) 21 39
Model for the 75th Percentile of 49 curves
Relative yield=exp(-(W126/65.49)**! .500) 15 24
"Based on composite functions for the 50th and 75th percentiles of 49 Weibull curves for relative above-ground biomass loss data
from 49 studies of 11 tree species grown under well-watered conditions for 1 or 2 year; curves were standardized to 90-day W126.
Source of parameters for the 49 curves: Lee and Hogsett (1996).
9.6.3 Validation of 1996 and 2006 Ozone AQCD Models and Methodology
Using the 90-day 12-h W126 and Current FACE Data
Since the completion of the NCLAN and NHEERL/WED projects, almost no studies
have been published that could provide a basis for estimates of exposure-response
that can be compared to those of the 1996 and 2006 O3 AQCDs. Most experiments,
regardless of exposure methodology, include only two levels of exposure.
In addition, very few studies have included measurements of exposure using the
W126 metric, or the hourly O3 concentration data that would allow computing
exposure using the W126. Two FACE projects, however, were conducted over
multiple years, and by adding to the number of exposure levels over time, can
support independent model estimation and prediction using the same model and the
same robust process as summarized in Section 9.6.2. Hourly O3 data were available
from both FACE projects.
The SoyFACE project is situated near Champaign, IL, and comprises 32 octagonal
rings (20m-diameter), 4 of which in a given year are exposed to ambient conditions,
and 4 of which are exposed to elevated O3 as a fixed proportion of the instantaneous
ambient concentration (Betzelberger et al.. 2010: UIUC. 2010: Morgan et al.. 2006:
Morgan et al.. 2004). Since 2002, yield data have been collected for up to 8
genotypes of soybean grown in subplots within each ring. The Aspen FACE project
is situated in Rhinelander, WI, and comprises 12 rings (30m-diameter), 3 of which
are exposed to ambient conditions, and 3 of which are exposed to O3 as a fixed
proportion of the instantaneous ambient concentration (Pregitzer et al.. 2008:
Karnosky et al.. 2005: Dickson et al.. 2000). In the summer of 1997, half the area of
each ring was planted with small (five to seven leaf sized) clonally propagated plants
of five genotypes of trembling aspen, which were left to grow in those environments
until 2009. Biomass data are currently available for the years 1997-2005 (King et al.,
2005). Ozone exposure in these two FACE projects can be viewed as a categorical
9-127
-------
variable with two levels: ambient, and elevated. However, this overlooks the facts
that not only do both ambient and elevated exposure vary from year to year, but the
proportionality between them also changes yearly. This change has two sources: first,
the dispensing of O3 into the elevated exposure rings varies from the set point for the
ambient/elevated proportionality to some extent, and for SoyFACE, the set point
changed between years. Second, when using threshold or concentration-weighted
cumulative metrics (such as AOT40, SUM06 or W126), the proportionality does not
propagate regularly from the hourly data to the yearly value. For example, hourly
average elevated exposures that are a constant 1.5 times greater than ambient do not
result in AOT40, SUM06 or W126 values that are some constant multiple of the
ambient values of those indices. Depending on the fraction of hourly values that are
above the threshold or heavily weighted, the same average yearly exposure will
result in different exposure values when using thresholded or weighted metrics.
In some years, elevated exposures in FACE experiments experience many more
values above the threshold, or more heavily weighted than the ambient exposures;
thus in those years, the distance between ambient and elevated exposure values
increases relative to other years. As a consequence, the number of exposure levels in
multi-year experiments is twice the number of years. In the case of SoyFACE for the
period between 2002 and 2008, ambient exposure in the highest year was
approximately equal to elevated exposure in the lowest year, with 14 levels of O3
exposure evenly distributed from lowest to highest. The particular conditions of the
Aspen FACE experiment resulted in 12 exposure levels between 1998 and 2003, but
they were not as evenly distributed between minimum and maximum over the 6-year
period.
There are necessary differences in the modeling of exposure-response in annual
plants such as soybean, and in perennial plants such as aspen trees, when exposure
takes place over multiple years. In annual plants, responses recorded at the end of the
life cycle, i.e., yearly, are analyzed in relationship to that year's exposure. Yield of
soybeans is affected by exposure during the year the crop was growing, and a new
crop is planted every year. Thus an exposure-response relationship can be modeled
from yearly responses matched to yearly exposures, with those exposure-response
data points having been generated in separate years. For perennial organisms, which
are not harvested yearly and continue to grow from year to year, such pairing of
exposure and response cannot be done without accounting for time. Not only does
the size of the organism at the beginning of each year of exposure increase, but size
is also dependent on the exposure from previous years. Therefore the relationship of
response and exposure must be analyzed either one year at a time, or by
standardizing the response as a yearly increment relative to size at the beginning of
each year. Furthermore, the relevant measurement of exposure is cumulative, or
cumulative yearly average exposure, starting in the year exposure was initiated, up to
the end of the year of interest. When analyzing the growth of trees over several years,
it would be evidently incorrect to pair the exposure level in every discrete year with
absolute size of the trees that year, and posit a direct relationship between them,
without taking increasing age into consideration. In the Aspen FACE experiment, for
example, one could not establish an exposure-response relationship by matching
12 yearly exposures and 12 yearly tree sizes, while disregarding age as if size did not
9-128
-------
also depend on it. This is the basis of the 2007 study of Aspen FACE data by Percy
et al. (2007), which compares the size of trees of various ages as if they were all the
same age, and was therefore not informative.
9.6.3.1 Comparison of NCLAN-Based Prediction and SoyFACE
Data
For this ISA, EPA conducted a comparison between yield of soybean as predicted by
the composite function three-step process (Section 9.6.2) using NCLAN data, and
observations of yield in SoyFACE. The median composite function for relative yield
was derived for the 11 NCLAN soybean Weibull functions for non-droughted
studies, and comparisons between the predictions of the median composite and
SoyFACE observations were conducted as follows.
For the years 2007 and 2008, SoyFACE yield data were available for 7 and 6
genotypes, respectively. The EPA used those data to compare the relative change in
yield observed in SoyFACE in a given year between ambient O3 and elevated O3,
versus the relative change in yield predicted by the NCLAN-based median composite
function between those same two values of O3 exposure. The two parameter median
composite function for relative yield of soybean based on NCLAN data was used to
predict yield response at the two values of exposure observed in SoyFACE in each
year, and the change between yield under ambient and elevated was compared to the
change observed in SoyFACE for the relevant year (Table 9-12). This approach
results in a direct comparison of predicted versus observed change in yield. Because
the value of relative response between any two values of O3 exposure is independent
of the intercept a, this comparison does not require prediction of the absolute values
of the responses.
Since comparisons of absolute values might be of interest, the predictive functions
were also scaled to the observed data: SoyFACE data were used to compute an
intercept a while the shape and scale parameters ((3 and r|) were held at their value in
the NCLAN predictive model. This method gives a comparison of prediction and
observation that takes all the observed information into account to provide the best
possible estimate of the intercept, and thus the best possible scaling (Table 9-13 and
Figure 9-18). For the comparison of NCLAN and SoyFACE, this validation was
possible for 2007 and 2008, where data for 7 and 6 soybean genotypes, respectively,
were available. The median composite function for relative yield was derived for the
11 NCLAN soybean Weibull functions for nondroughted studies, and the values of
median yield under ambient exposure at SoyFACE in 2007 and 2008 were used to
obtain an estimate of the intercept a for the NCLAN median function in each of the
two years. Table 9-12 presents the results of ambient/elevated relative yield
comparisons between the NCLAN-derived predictions and SoyFACE observations.
Table 9-13 and Figure 9-18 present the results of comparisons between NCLAN-
derived predictions and SoyFACE observations of yield, with the predictive function
scaled to provide absolute yield values.
9-129
-------
Table 9-12 Comparison between change in yield observed in the SoyFACE
experiment between elevated and ambient O3, and change
predicted at the same values of O3 by the median composite
function for NCLAN.
Year
2007
2008
90-day 12-h
observed
Ambient
4.39
3.23
W126 (ppm-h)
at SoyFACE
Elevated
46.23
28.79
Yield in Elevated O3
Predicted by
NCLAN3
75
85
Relative to Ambient O3 (%)
Observed at SoyFACE
76
88
"Two-parameter relative yield model.
Table 9-13 Comparison between yield observed in the SoyFACE experiment
and yield predicted at the same values of O3 by the median
composite function for NCLAN.
90-day 12-h W126 (ppm-h)
observed at SoyFACE
Year
2007
2008
Ambient
4.39
3.23
Elevated
46.23
28.79
Yield predicted by
NCLAN3 (g/m2)
Ambient
309.2
350.3
Elevated
230.6
298.2
Yield observed at
SoyFACE (g/m2)
Ambient
305.2
344.8
Elevated
230.6
304.4
"Three-parameter absolute yield model with intercept scaled to SoyFACE data.
9-130
-------
400
350
300
250
200
150
100
50
0
2007, 7 genotypes
20 30 40 50 60 70
90day12hrW126 (ppm-hr)
400
350
300
_ 250
S 200
•V
* 150 -
100 -
50 -
0
2008, 6 genotypes
20 30 40 50
90 day 12 hr W126 (ppm-hr)
Note: Black dots are the median of 7 or 6 soybean genotypes in SoyFACE (2007, 2008); bars are Inter-Quartile Range for
genotypes; dashed line is median composite model for 11 studies in NCLAN.
Source of data: Betzelberger et al. (2010). Morgan et al. (2006). Lee and Hogsett (1996).
Figure 9-18 Comparison of yield observed in SoyFACE experiment in a given
year with yield predicted by the median composite function based
on NCLAN.
Finally, a composite function for the 25th, 50th, and 75th percentiles was developed
from SoyFACE annual yield data, and compared to the NCLAN-based function.
The process described in Section 9.6.2 was applied to SoyFACE data for individual
genotypes, aggregated over the years during which each was grown; one genotype
from 2003 to 2007, and six genotypes in 2007 and 2008. First, the three parameter
Weibull model described in Section 9.6.2 was estimated using nonlinear regression
on exposure-yield data for each genotype separately, over the years for which data
were available, totaling seven curves. The 25th, 50th, and 75th percentiles of the
predicted values for the two parameter relative yield curves were then identified at
every integer of W126 between 0 and 60, and a two-parameter Weibull model
estimated by regression for the three quartiles. The comparison between these
composite functions for the quartiles of relative yield loss in SoyFACE and the
corresponding composite functions for NCLAN is presented in Figure 9-19.
9-131
-------
100 -
90 -
80 -
g 70 -
1 60 H
1 50 •
I 40 -
-------
9.6.3.2 Comparison of NHEERL/WED-Based Prediction of Tree
Biomass Response and Aspen FACE Data
EPA also conducted two comparisons between prediction of above-ground biomass
loss based on NHEERL/WED results and observations from Aspen FACE.
The median composite function was developed from NHEERL/WED data for 11
studies that used wild-type seedlings of aspen as well as four clonally propagated
genotypes. All plants were grown in OTCs for one growing season before being
destructively harvested. Aspen FACE data were from clonally propagated trees of
five genotypes grown from 1998 to 2003, with above-ground biomass calculated
using allometric equations derived from data for trees harvested destructively in 2000
and 2002 (King et al. 2005).
The two parameter median composite function for relative biomass was used to
predict biomass response under the observed elevated exposure, relative to its value
under observed ambient exposure, for each separate year of Aspen FACE. EPA first
compared Aspen FACE observations of the change in biomass between ambient and
elevated exposure with the corresponding prediction at the same values of exposure.
Comparisons between observed and predicted absolute biomass values were then
conducted for each year by scaling the predictive function to yearly Aspen FACE
data as described for soybean data in Section 9.6.3.1. In all cases, yearly 90 day
12-hour W126 values for Aspen FACE were computed as the cumulative average
from the year of planting up to the year of interest. A comparison of composite
functions between NHEERL/WED and Aspen FACE, similar to the one performed
for NCLAN and SoyFACE, was not possible: as discussed in the introduction to
Section 9.6, the pairing of 12 exposure values from separate years and 12 values of
biomass cannot be the basis for a model of exposure-response, because the trees
continued growing for the six-year period of exposure. Because the same trees were
used for the entire duration, and continued to grow, data could not be aggregated
over years. Table 9-14 presents the results of ambient/elevated relative biomass
comparisons between the NHEERL/WED-derived predictions and Aspen FACE
observations. Table 9-15 and Figure 9-20 present the results of the comparison
between NHEERLAVED-derived predictions and Aspen FACE observations for
absolute biomass, using Aspen FACE data to scale the NHEERL/WED-derived
composite function.
9-133
-------
Table 9-14 Comparison between change in
and ambient O3 in Aspen FACE
predicted at the same values of
function for NHEERL/WED.
above-ground biomass elevated
experiment in 6 year, and change
O3 by the median composite
90-day 12-h W126 (ppm-h)
Cumulative Average observed at Aspen FACE
Year
1998
1999
2000
2001
2002
2003
Ambient
3.19
2.61
2.43
2.55
2.51
2.86
Elevated
30.08
33.85
30.16
31.00
30.27
29.12
Above-Ground Biomass in Elevated O3,
Relative to Ambient O3 (%)
Predicted by
NHEERL/WED3
74
70
74
73
74
75
Observed at
Aspen FACE
75
70
71
71
69
71
"Two-parameter relative biomass model
Table 9-15 Comparison between above-ground biomass observed in Aspen
FACE experiment in 6 year and biomass predicted by the median
composite function based on NHEERL/WED.
90 day 12-h W126 (ppm-h)
Cumulative Average observed at
Aspen FACE
Year
1998
1999
2000
2001
2002
2003
Ambient
3.19
2.61
2.43
2.55
2.51
2.86
Elevated
30.08
33.85
30.16
31.00
30.27
29.12
Biomass Predicted by
NHEERL/WEDa (g/m2)
Ambient
276.0
958.7
1,382.4
1,607.0
2,079.0
2,640.1
Elevated
203.2
668.3
1 ,022.8
1,173.7
1,532.1
1,981.2
Biomass Observed at
Aspen FACE (g/m2)
Ambient
274.7
955.3
1 ,400.3
1 ,620.7
2,125.9
2,695.2
Elevated
204.9
673.3
998.6
1,154.9
1,468.4
1,907.8
"Three-parameter absolute biomass model with intercept scaled to Aspen FACE data.
9-134
-------
(M
Biomass (g
3000 -|
2500 -
2000 -
1500 -
1000 -
500 -
n .
^
^
"""•-. „ 1 2003
-»--- """-I
-*_ """"--^ I 2°°2
*~~*--I"i 2001
_ ^ "•» 2000
""""""•-« 1999
-f 1998
10 20 30 40 50 60
90 day 12 hr W126 (yearly cumulative average, ppm-hr)
70
Note: Black dots are aspen biomass/m for 3 FACE rings filled with an assemblage of 5 clonal genotypes of aspen at Aspen FACE;
bars are SE for 3 rings; dashed line is median composite model for 4 clonal genotypes and wild-type seedlings in 11
NHEERL/WED 1 -year OTC studies.
Source of data: King et al. (2005). Lee and Hogsett (1996).
Figure 9-20 Comparison between above-ground biomass observed in Aspen
FACE experiment in 6 year and biomass predicted by the median
composite function based on NHEERL/WED.
As in the comparisons between NCLAN and SoyFACE, the agreement between
predictions based on NHEERL/WED data and Aspen FACE observations was very
close. The results of the two projects strongly reinforce each other with respect to the
response of aspen biomass to O3 exposure. The methodology used for obtaining the
median composite function is shown to be capable of deriving a predictive model
despite potential confounders, and despite the added measurement error that is
expected from calculating biomass using allometric equations. In addition, the
function based on one year of growth was shown to be applicable to subsequent
years.
The results of experiments that used different exposure methodologies, different
genotypes, locations, and durations converged to the same values of response to O3
exposure for each of two very dissimilar plant species, and predictions based on the
earlier experiments were validated by the data from current ones. However, in these
comparisons, the process used in establishing predictive functions involved
aggregating data over variables such as time, locations, and genotypes, and the use of
a robust statistic (quartiles) for that aggregation. The validating data, from SoyFACE
and Aspen FACE, were in turn aggregated over the same variables. The accuracy of
9-135
-------
predictions is not expected to be conserved for individual values of those variables
over which aggregation occurred. For example, the predicted values for soybean,
based on data for five genotypes, are not expected to be valid for each genotype
separately. As shown in the validation, however, aggregation that occurred over
different values of the same variable did not affect accuracy: composite functions
based on one set of genotypes were predictive for another set, as long as medians
were used for both sets. A study of cottonwood (Populus deltoides) conducted using
a naturally occurring gradient of O3 exposure (Gregg et al. 2006. 2003) may provide
an illustration of the response of an individual species whose response is far from the
median response for an aggregation of species.
9.6.3.3 Exposure-Response in a Gradient Study
Gregg et al. (2003) grew saplings of one clonally propagated genotype of
cottonwood (Populus deltoides} in seven locations within New York City and in the
surrounding region between July and September in 1992, 1993 and 1994, and
harvested them 72 days after planting. Owing to regional gradients of atmospheric
O3 concentration, the experiment yielded eight levels of exposure (Figure 9-21), and
the authors were able to rule out environmental variables other than O3 to account for
the large differences in biomass observed after one season of growth. The deficit in
growth increased substantially faster with increasing O3 exposure than has been
observed in aspen, another species of the same genus (Populus tremuloides, Section
9.6.3.2). Using a three parameter Weibull model (Figure 9-21), the biomass of
cottonwood at a W126 exposure of 15 ppm-h, relative to biomass at 5 ppm-h, is
estimated to be 0.18 (18% of growth at 5 ppm-h). The relative biomass of trembling
aspen within the same 5-15 ppm-h range of exposure is estimated to be 0.92, using
the median composite model for aspen whose very close agreement with Aspen
FACE data was shown in Section 9.6.3.2. Using a median composite function for all
deciduous trees in the NHEERL/WED project (6 species in 21 studies) also gives
predictions that are very distant from the cottonwood response observed in this
experiment. For all deciduous tree species in NHEERLAVED, biomass at a W126
exposure of 15 ppm-h, relative to biomass at 5 ppm-h, was estimated to be 0.87.
9-136
-------
100 -
90 -
80 -
70 -
o> 60 -
| 50 H
m 40 -
30 -
20 -
10 -
0
0 10 20 30 40 50
72 day 12hrW126 (ppm-hr)
60
70
Note: Line represents the three-parameter Weibull model.
Source: Modified with permission of Nature Publishing Group (Gregg et al.. 2003).
Figure 9-21 Above-ground biomass for one genotype of cottonwood grown in
seven locations for one season in 3 years.
As shown in Section 9.6.2, the median models available for trembling aspen and
soybean have verifiable predictive ability for those particular species. This suggests
that the corresponding NCLAN- and NHEERL/WED-based models for multiple crop
and tree species can provide reliable estimates of losses for similar assortments of
species. However, their predictive ability would likely be poor for individual species
not tested.
The cottonwood data of Gregg et al. (2003) show an extremely severe response to
O3. They are consistent with the expectation that among species and genotypes, some
are likely to be substantially more sensitive than a median measure, such as the
estimate produced by NHEERL/WED (Figure 9-16), but the sensitivity of this
particular species has not been studied elsewhere.
An alternative hypothesis for the difference between the response of cottonwood in
this experiment and deciduous tree species in NHEERL/WED, or the difference
between the response of cottonwood and aspen in NHEERL/WED and Aspen FACE,
could be the presence of confounding factors in the environments where the
experiment was conducted. However, variability in temperature, moisture, soil
fertility, and atmospheric deposition of N were all ruled out by Gregg et al. (2003) as
contributing to the observed response to O3. In addition, this hypothesis would imply
9-137
-------
that the unrecognized confounder(s) were either absent from both OTC and FACE
studies, or had the same value in both. This is not impossible, but the hypothesis that
cottonwood is very sensitive to O3 exposure is more parsimonious, and sufficient.
9.6.3.4 Meta-analyses of Growth and Yield Studies
Since the 2006 O3 AQCD, five studies have used meta-analytic methods to integrate
results from experimental studies of crops or tree species relevant to the United
States. It is possible to obtain exposure-response data for growth and yield from
those meta-analyses, but because all of them provided summary measurements of O3
exposure as hourly averages of various lengths of exposures, comparisons with
exposure-response results where exposure is expressed as W126 are problematic.
Table 9-16 summarizes the characteristics of the five meta-analyses. They all
included studies conducted in the U.S. and other locations worldwide, and all of them
expressed responses as comparative change between levels of exposure to O3, with
carbon filtered air (CF) among those levels. Using hourly average concentration to
summarize exposure, CF rarely equates with absence of O3, although it almost
always near zero when exposure is summarized as W126, SUM06, or AOT40.
Table 9-16 Meta-analyses of growth or yield studies published since 2005.
Study
Ainsworth
(2008)
Feng et al.
(2008b)
Feng and
Kobayashi
(2009)
Grantz et al.
(2006)
Wittig et al.
(2009)
Number of
articles included
12
53
All crops together: 81
16
All responses:263
Articles that included
biomass:unreported
Years of
publication
surveyed
1 980-2007
1980-2007
1980-2007
1 992-2004
1 970-2006
Crop, species
or genera
Rice
Wheat
Potato, barley, wheat,
rice, bean, soybean
34 Herbaceous dicots
21 Herbaceous monocots
5 Tree species
4 Gymnosperm tree
genera
1 1 Angiosperm tree
genera
Response
Yield
Yield
Yield
Relative
Growth Rate
Total
biomass
Number
ofO3
levels
2
5
3
2
4
Duration of
exposure
unreported
>10 days
>10 days
2-24 weeks
>7 days
9-138
-------
The only effect of O3 exposure on yield of rice reported in Ainsworth (2008) was a
decrease of 14% with exposure increasing from CF to 62 ppb average concentration.
Feng et al. (2008b) were able to separate exposure of wheat into four classes with
average concentrations of 42, 69, 97, and 153 ppb, in data where O3 was the only
treatment. Mean responses relative to CF were yield decreases of 17, 25, 49, and
61% respectively. Feng et al. (2008b) observed that wheat yield losses were smaller
under conditions of drought, and that Spring wheat and Winter wheat appeared
similarly affected. However, mean exposure in studies of Winter wheat was
substantially higher than in studies of Spring wheat (86 versus 64 ppb), which
suggests that the yield of Spring wheat was in fact more severely affected, since yield
was approximately the same, even though Spring wheat was exposed to lower
concentrations. Exposures of the six crops considered in Feng and Kobayahi (2009)
were classified into two ranges, each compared to CF air. In the lower range of
exposure (41-49 ppb), potato studies had the highest average exposure (45 ppb) and
wheat and rice the lowest (41 ppb). In the higher range (51-75 ppb), wheat studies
had the highest average exposure (65 ppb), and potato, barley and rice the lowest
(63 ppb). In other words, across the studies included, all crops were exposed to very
similar levels of O3. At approximately 42 ppb, the yield of potato, barley, wheat,
rice, bean, and soybean declined by 5.3, 8.9, 9.7, 17.5, 19, and 7.7% respectively,
relative to CF air. At approximately 64 ppb O3, declines were 11.9, 12.5, 21.1, 37.5,
41.4, and 21.6%. Grantz et al. (2006) reported Relative Growth Rate (RGR) rather
than growth, and did not report O3 exposure values in a way that would allow
calculation of mean exposure for each of the three categories of plants for which
RGR changes are reported. All studies used only two levels of exposure, with CF air
as the lower one, and most used elevated exposure in the range of 40 to 70 ppb.
Decline in RGR was 8.2% for the 34 herbaceous dicots, 4.5% for the 21 herbaceous
monocots, and 17.9% for the 5 tree species. Finally, Wittig et al. (2009) divided the
studies analyzed into three classes of comparisons: CF versus ambient, CF versus
elevated, and ambient versus elevated, but reported comparisons between three
average levels of exposure besides CF: 40 ppb, 64 ppb, and 97 ppb. Corresponding
decreases in total biomass relative to CF were 7, 17, and 17%.
These meta-analyses provide very strong confirmation of EPA's conclusions from
previous O3 AQCDs: compared to lower levels of ambient O3, current levels in many
locations are having a substantial detrimental effect on the growth and yield of a
wide variety of crops and natural vegetation. They also confirm strongly that
decreases in growth and yield continue at exposure levels higher than current
ambient levels. However, direct comparisons with the predictions of exposure-
response models that use concentration-weighted cumulative metrics are difficult.
9-139
-------
9.6.3.5 Additional Exposure-Response Data
The studies summarized in Table 9-17 and Table 9-18 contain growth or yield
exposure-response data at too few levels of exposure for exposure-response models,
and/or used metrics other than W126. These tables update Tables AX9-16 through
AX9-19 of the 2006 O3 AQCD.
9.6.4 Summary
None of the information on effects of O3 on vegetation published since the 2006 O3
AQCD has modified the assessment of quantitative exposure-response relationships
that was presented in that document. This assessment updates the 2006 exposure-
response models by computing them using the W126 metric, cumulated over
90 days. Almost all of the experimental research on the effects of O3 on growth or
yield of plants published since 2006 used only two levels of exposure. In addition,
hourly O3 concentration data that would allow calculations of exposure using the
W126 metric are generally unavailable. However, two long-term experiments, one
with a crop species (soybean), one with a tree species (aspen), have produced data
that can be used to validate the exposure-response models presented in the 2006 O3
AQCD, and methodology used to derive them.
Quantitative characterization of exposure-response in the 2006 O3 AQCD was based
on experimental data generated for that purpose by the National Crop Loss
Assessment Network (NCLAN) and EPA National Health and Environmental Effects
Research Laboratory, Western Ecology Division (NHEERL-WED) projects, using
OTCs to expose crops and trees seedling to O3. In recent years, yield and growth
results for two of the species that had provided extensive exposure-response
information in those projects have become available from studies that used FACE
technology, which is intended to provide conditions much closer to natural
environments (Pregitzer et al.. 2008: Morgan et al.. 2006: Morgan et al.. 2004:
Dickson et al.. 2000). The robust methods that were used previously with exposure
measured as SUM06 were applied to the NCLAN and NHEERL-WED data with
exposure measured as W126, in order to derive single-species median models for
soybean and aspen from studies involving different genotypes, years, and locations.
The resulting models were used to predict the change in yield of soybean and
biomass of aspen between the two levels of exposure reported in recent FACE
experiments. Results from these new experiments were exceptionally close to
predictions from the models. The accuracy of model predictions for two widely
different plant species provides support for the validity of the corresponding
multiple-species models for crops and trees in the NCLAN and NHEERL-WED
projects. However, variability among species in those projects indicates that the
range of sensitivity is likely quite wide. This was confirmed by a recent experiment
with cottonwood in a naturally occurring gradient of exposure (Gregg et al.. 2006).
which established the occurrence of species with responses substantially more severe
9-140
-------
under currently existing conditions than are predicted by the median model for
multiple species.
Results from several meta-analyses have provided approximate values for responses
of yield of soybean, wheat, rice and other crops under broad categories of exposure,
relative to char coal-filtered air (Ainsworth. 2008: Feng et al.. 2008b: Morgan et al..
2003). Likewise, Feng and Kobayashi (2009) have summarized yield data for six
crop species under various broad comparative exposure categories, while Wittig et al.
(2009) reviewed 263 studies that reported effects on tree biomass. However, these
analyses have proved difficult to compare with exposure-response models, especially
given that exposure was not expressed in the same W126 metric.
Table 9-17 Summary of studies of effects of O3 exposure on growth and yield
of agricultural crops.
Species
Facility Exposure
Location Duration
Alfalfa (Medicago 2 yr, 2005,
sativa) 2006
OTC; 0.27m3 pots
Federico, Italy
Bean (Phaseolus 3 months,
vulgaris I. cv 2006
Borlotto)
OTC; ground-
planted
Curno, Italy
Big Blue Stem 4 months,
(Andropogon 2003
gerardii)
OTC
Alabama
Brassica napus 1 7-26 days
cv. We star
Growth chambers
Finland
Corn (Zea mays 33 days
cv. Chambord)
OTC
France
Cotton cv. Pima 8 weeks
OTC; 9-L pots
San Joaquin
Valley, CA
O3 Exposure
(Additional Treatment)
AOT40: CF 0 ppm-h
13.9ppm-h (2005),
10.1 ppm-h (2006)
(NaCI: 0.29, 0.65, 0.83,
1 .06 deciSiemens/meter)
Seasonal AOT40:
CF (0.5 ppm-h);
ambient (4.6 ppm-h)
(N/A)
12-havg:
CF(14ppb),
Ambient (29 ppb),
Elevated (71 ppb)
(N/A)
8-h avg:
CF(Oppb), 100 ppb
(Bt/non-Bt; herbivory)
AOT40 ppm-h: 1.1; 1.3; 4.9;
7.2; 9.3; 12.8
(N/A)
12-havg: 12.8 ±0.6;
79.9 ±6.3; 122.7 ±9.7
(N/A)
Response
Measured
Total shoot yield
# Seeds per plant;
100-seed weight
Final harvest
biomass;
RVF
Shoot biomass
Total above-
ground biomass
Above-ground
biomass
Percent Change
from CF
(Percent Change
from Ambient) Reference
n.s. (N/A) Maggio et al.
(2009)
-33 (N/A) Gerosa et al.
n.s. (N/A) (2009)
n.s. (n.s.) Lewis etal.
.7 (-7) (2QQ6)
-30.70 (N/A) Himanen et al.
(2009b)
N/A (Highest Leitao et al.
treatment caused - (2007a)
26% change)
-76 (n.s.) Grantz and
Shrestha (2006)
9-141
-------
Species
Facility
Location
Eastern
Gamagrass
(Tripsacum
dactyloides)
OTC
Alabama
Grapevine (Vitis
vinivera)
OTC
Austria
Mustard (Brassica
campestris)
Chambers;
7.5-cm pots
Oilseed Rape
(Brassica napus)
OTC
Yangtze Delta,
China
Peanut (Arachis
hypogaea)
OTC
Raleigh, NC
Poa pratensis
OTC
Braunschweig,
Germany
Potato (Solanum
tuberosum)
OTC; CHIP
6 northern
European
locations
Rice (Oryza
sativa)
OTC
Raleigh, NC
Exposure
Duration
4 months,
2003
3 yr, May-Oct
1 0 days
39 days
Syr
2000-2002:
4-5 weeks in
the Spring
1988,1999.
Emergence to
harvest
1997-1998,
June-
September
Os Exposure
(Additional Treatment)
12-h avg:
CF(14ppb),
Ambient (29 ppb),
Elevated (71 ppb)
(N/A)
AOT40 ppm-h:
CF (0),
Ambient (7-20),
Elevated. 1 (20-30),
Elevated. 2 (38-48)
CF&
67.8 ppb for 7 h
(N/A)
Daily avg: 100 ppb, one with
diurnal variation and one
with constant concentration
(N/A)
12-h avg:
CF (22 ppb),
Ambient (46 ppb),
Elevated (75ppb)
(CO2: 375 ppm; 548 ppm;
730 ppm)
8-h avg:
CF+25 (21 .7),
NF+50(73.1)
(Competition)
AOT40:CF (0);
Ambient (0.27-5. 19); NF
(0.002-2.93)
NF+ (3.10-24.78
(N/A)
12-h mean ppb:
CF (27.5),
Elevated (74.8)
(C02)
Response
Measured
Final harvest
biomass;
RVF
Total fruit yield/
Sugar yield
Seeds/plant
Biomass and
pods per plant
Yield (seed
weight, g/m)
Total biomass (g
DW/pot)
Tuber yield
averaged across
5 field-sites;
Tuber starch
content regressed
against [O3]
report sig. ± slope
with increasing
[03]
Total biomass;
Seed yield
Percent Change
from CF
(Percent Change
from Ambient)
+68 (+42);
-17 (-12)
-20 to -80 in
different yr
(-20 to -90 in
different yr)
n.s. (N/A)
Diurnal variability
reduced both
biomass and pod
number more than
constant fumigation
(N/A)
-33 (-8)
N/A (n.s.)
N/A (-27% -+27%,
most comparisons
n.s.) Linear
regression
slope = -0.0098)
-25(N/A)
-13 to 20 (N/A)
Reference
Lewis et al.
(2006)
Soja et al.
(2004)
Black et al.
(2007)
Wang et al.
(2008)
Burkey et al.
(2007)
Bender et al.
(2006)
Vandermeiren
et al. (2005)
Reid and Fiscus
(2008)
9-142
-------
Species
Facility
Location
Rice (Oryza
sativa) 20 Asian
cultivars
OTC
Gunma
Prefecture, Japan
Seminatural grass
FACE
Le Mouret,
Switzerland
Soybean
OTC; CRA
Bari, Italy
Soybean (Glycine
maxcv. 93B15)
SoyFACE
Urbana, IL
Soybean (Glycine
maxcv. Essex)
Chambers; 21 L
Raleigh, NC
Soybean (Glycine
maxcv. Essex)
OTCs;21-Lpots
Raleigh, NC
Soybean (Glycine
maxcv. Tracaja)
Chambers; pots
Brazil
Soybean (Glycine
max) 10 cultivars
SoyFACE
Urbana, IL
Spring Wheat
(Triticum aestivum
cv. Minaret; Satu;
Drabant; Dragon)
OTCs
Belgium, Finland,
& Sweden
Strawberry
(Fragaria x
ananassa Duch.
Cv Korona &
Elsanta)
Growth chambers
Bonn, Germany
Exposure
Duration
2008 growing
season
Syr
2003-2005
growing
seasons
2002, 2003
growing
seasons
2x3 months
2x3 months
20 days
2007 & 2008
1990-2006
2 months
Os Exposure
(Additional Treatment)
Daily avg (ppb):
CF (2),
O.Sxambient (23);
1 xambient (28);
1.5xambient (42);
2xambient (57)
(Cultivar comparisons)
Seasonal AOT40: Ambient
(0.1-7.2ppm-h);
Elevated. (1.8-24.1 ppm-h)
(N/A)
Seasonal AOT40 ppm-h: CF
(0),
Ambient (3.4), High (9.0)
(Drought)
8-h avg:
Ambient (62 & 50 ppb),
Elevated (75 & 63 ppb)
(N/A)
12-havg:CF(28),
Elevated (79),
Elevated flux (1 1 2)
(CO2: 365 & 700)
12-havg:CF(18);
Elevated (72)
(C02:367&718)
12-h avg: CF & 30 ppb
(N/A)
8-h avg: Ambient (46.3 &
37.9), Elevated (82.5 & 61 .3)
(Cultivar comparisons)
Seasonal AOT40s ranged
from 0 to16 ppm-h
(N/A)
8-h avg: CF (0 ppb) &
Elevated (78 ppb)
(N/A)
Response
Measured
Yield
Relative annual
yield
Yield
Yield
Seed mass per
plant
Seed mass per
plant
Biomass
Yield
Seed protein
content;
1 ,000-seed weight
regressed across
all experiments
Fruit yield
(weight/plant)
Percent Change
from CF
(Percent Change
from Ambient)
From n.s. to -30
across all cultivars
N/A (2xfaster
decrease in yield/yr)
-46 (-9)
N/A
(-15 in 2002;
-25 in 2003)
-30 (N/A)
-34 (N/A)
-18 (N/A)
N/A (-17.20)
N/A (significant
negative correlation)
N/A (sig negative
correlation)
-1 6 (N/A)
Reference
Sawada and
Kohno (2009)
Volketal.
(2006)
Bou Jaoude
et al. (2008b)
Morgan et al.
(2006)
Booker and
Fiscus (2005)
Booker et al.
(2004b)
Bulbovas et al.
(2007)
Betzelberger et
al. (2010)
Piikki et al.
(2008b)
Keutgen et al.
(2005)
9-143
-------
Species
Facility
Location
Sugarbeet (Beta
vulgaris cv.
Patriot)
OTC
Belgium
Sugarcane
(Saccharum spp)
CSTR
San Joaquin
Valley, CA
Sweet Potato
Growth chambers
Bonn, Germany
Tomato
(Lycopersicon
esculentum)
OTC
Valencia, Spain
Trifolium
Subterraneum
OTC; 2.5-L pots
Madrid, Spain
Watermelon
(Citrullus lanatus)
OTC
Valencia, Spain
Yellow Nutsedge
OTC; 9-L pots
San Joaquin
Valley, CA
Exposure
Duration
2003, 2004;
5 months
2007;
11 -13 weeks.
4 weeks
1 33 days in
1998
29 days
2000, 2001 .
90 days
8 weeks
Os Exposure
(Additional Treatment)
8-h avg: Ambient (36 ppb);
Elevated (62 ppb)
(N/A)
12-havg: CF (4 ppb);
Ambient (58);
Elevated (147)
(N/A)
8-h avg: CF (0 ppb),
Ambient (<40 ppb) Elevated
(255 ppb)
(N/A)
8-h mean ppb:
CF16.3, NF30.1,
NF+ 83.2
(Various cultivars; early &
late harvest)
12-havg:CF(<7.9±6.3);
Ambient (34.4 ± 10.8);
Elevated (56.4 ± 22.3)
(N: 5, 15 & 30 kg/ha)
AOT40: CF (0 ppm-h)
Ambient (5.7 ppm-h),
Elevated (34.1 ppm-h)
(N:0, 1 4.0 & 29.6 g/pot)
12-havg: 12.8 ±0.6;
79.9 ±6.3; 122.7 ±9.7
(N/A)
Percent Change
from CF
Response (Percent Change
Measured from Ambient)
Sugar yield N/A (-9)
Total biomass -40 (-30)
(g/plant)
Tuber weight -14 (-11. 5)
Yield n.s(n.s.)
Above-ground -45 (-35)
biomass
total fruit yield (kg) n.s. (54)
above-ground n.s. (n.s.)
biomass
Reference
De Temmerman
et al. (2007)
Grantz and Vu
(2009)
Keutgen et al.
(2008)
Dalstein and
Vas (2005)
Sanz et al.
(2005)
Calatayud et al.
(2006)
Grantz and
Shrestha (2006)
In studies where variables other than O3 were included in the experimental design, response to O3 is only provided for the
control level of those variables.
9-144
-------
Table 9-1 8 Summary of
vegetation.
Species
Facility
Location
Yellow nutsedge (Cyperus
esculentus)
CSTR
San Joaquin Valley, CA
35 herbaceous species
OTC
Corvallis, OR
Highbush blackberry
(Rubus argutus)
OTC
Auburn, AL
Horseweed (Conyza
canadensis)
CSTR
San Joaquin Valley, CA
Red Oak (Quercus
rubrum)
Forest sites
Look Rock & Twin Creeks
Forests, TN
Pine species
Forest sites
Look Rock Forest, TN
Exposure
Duration
53 days in
2008
1999-2002,
May-August
2004,
May-August
2005, 2 runs,
28 days each
(July-Aug,
Sept)
2001 -2003,
April-October
2001 -2003,
April-October
studies of effects of O3 exposure on growth of natural
O3 Exposure
(Additional
Treatment)
12-h mean ppb:
CF (4); CF+ (60);
CF2+(115)
4-yr avg; yearly
W126ppm-h:
CF (0),
CF+ (21),
CF 2+ (49.5)
12-h mean ppb:
CF (21.7),
Ambient (32.3),
Elevated (73.3)
W126ppm-h:
CF(0),
CF 2+ (30)
(Glyphosate
resistance)
AOT60:
2001 (11.5),
2002 (24.0),
2003 (1 1 .7)
(Observational
study with
multiple
environmental
variables)
AOT60:
2001 (11.5),
2002 (24.0),
2003 (1 1 .7)
(Observational
study with
multiple
environmental
variables)
Response
Measured
Above-ground
biomass; tubers
(g/plant)
Total community
above-ground
biomass (35 species)
after 4 years
Vegetative regrowth
after pruning
Total biomass
(g/plant)
Annual circumference
increment (change
relative to 2001 in
year 2002;2003)
Annual circumference
increment (change
relative to 2001 in
year 2002;2003)
Response
ns; CF(4.1)
CF+(3.9) CF2+(2.7)
CF (459 g/m2), CF+
(457 g/rri ), CF2+
(398 g/m2)
CF (75.1 g/plant),
Ambient (76.4
g/plant),
Elevated (73.1
g/plant)
Glyphosate
sensitive:
CF (0.354)
CF+(0.197)
CF2+ (0.106)
Glyphosate
resistant: CF(0.510)
CF+ (0.313)
CF2+(0.143)
-42.8%; +1%
-62.5%; -2.9%
Reference
Grantz et al. (201 Oa)
Pfleegeretal. (201 0)
Ditchkoffetal. (2009)
Grantz et al. (2008)
McLaughlin et al.
(2QQ7a)
McLaughlin et al.
(2QQ7a)
9-145
-------
Species
Facility
Location
Hickory species
Forest sites
Look Rock Forest, TN
Chestnut Oak (Quercus
prinus)
Forest sites
Look Rock Forest, TN
Black Cherry (Prunus
rigida)
Forest sites
Twin Creeks Forest, TN
Shortleaf pine (Pinus
echinata)
Forest sites
Twin Creeks Forest, TN
Hemlock (Tsuga
canadensis)
Forest sites
Twin Creeks Forest, TN
Red Maple (Acer rubrum)
Forest sites
Twin Creeks Forest, TN
Exposure
Duration
2001 -2003,
April-October
2001 -2003,
April-October
2002-2003,
April-October
2002-2003,
April-October
2002-2003,
April-October
2002-2003,
April-October
O3 Exposure
(Additional
Treatment)
AOT60:
2001 (11.5),
2002 (24.0),
2003 (1 1 .7)
(Observational
study with
multiple
environmental
variables)
AOT60:
2001 (11.5),
2002 (24.0),
2003 (1 1 .7)
(Observational
study with
multiple
environmental
variables)
AOT60:
2002 (24.0),
2003 (1 1 .7)
(Observational
study with
multiple
environmental
variables)
AOT60:
2002 (24.0),
2003 (1 1 .7)
(Observational
study with
multiple
environmental
variables)
AOT60:
2002 (24.0),
2003 (1 1 .7)
(Observational
study with
multiple
environmental
variables)
AOT60:
2002 (24.0),
2003 (1 1 .7)
(Observational
study with
multiple
environmental
variables)
Response
Measured Response
Annual circumference -14%; +30%
increment (change
relative to 2001 in
year 2002;2003)
Annual circumference +44%; +55%
increment (change
relative to 2001 in
year 2002;2003)
Annual circumference -75%
increment (change
relative to 2003 in
year 2002)
Annual circumference -16.8%
increment (change
relative to 2003 in
year 2002)
Annual circumference -21 .9%
increment (change
relative to 2003 in
year 2002)
Annual circumference -59.6%
increment (change
relative to 2003 in
year 2002)
Reference
McLaughlin et al.
(2007a)
McLaughlin et al.
(2007a)
McLaughlin et al.
(2QQ7a)
McLaughlin et al.
(2007a)
McLaughlin et al.
(2QQ7a)
McLaughlin et al.
(2007a)
9-146
-------
Species
Facility
Location
Yellow Poplar
(Liriodendron tulipifera)
Forest sites
Look Rock, Oak Ridge, &
Twin Creeks Forest, TN
Sugar Maple (Acer
saccharum)
Forest sites
Twin Creeks Forest, TN
Trembling aspen (Populus
tremuloides), 5 genotypes
Aspen FACE
Rhinelander, Wl
Hybrid Poplar (Populus
trichocarpa x Populus
deltoides)
OTC
Seattle, WA
Exposure
Duration
2002-2003,
April-October
2002-2003,
April-October
1998-2004,
May-October
2003,
3 months
O3 Exposure
(Additional
Treatment)
AOT60:
2002 (24.0),
2003 (1 1 .7)
(Observational
study with
multiple
environmental
variables)
AOT60:
2002 (24.0),
2003 (1 1 .7)
(Observational
study with
multiple
environmental
variables)
Cumulative avg
90-day 1 2-h
W126.
Ambient
3.1 ppm-h
Elevated:
27.2 ppm-h
(Competition with
birch, maple)
Daily mean
(ug/g):
CF(<9),
Elevated (85-
128)
Response
Measured Response
Annual circumference -45.9%; -15.25%
increment (change
relative to 2001 in
years 2002; 2003)
Annual circumference -63.8%
increment (change
relative to 2003 in
year 2002)
main stem volume Ambient: 6.22 dm3;
after 7 years Elevated: 4.73 dm3
Total biomass CF to elevated:
-12.9%
Reference
Mclaughlin et al.
(2007a)
Mclaughlin et al.
(2007a)
Kubiske et al. (2006)
Woo and Hinckley
(2005)
In studies where variables other than O3 were included in the experimental design, response to O3 is only provided for the control
level of those variables.
9.7 Summary and Conclusions
Based on the evidence presented in Chapter 9 and summarized here, O3 is causally
related or likely to be causally related to effects observed on vegetation and
ecosystems. The evidence for these effects spans the entire continuum of biological
organization, from the cellular and subcellular level to the whole plant, and up to
ecosystem-level processes, and includes evidence for effects at lower levels of
organization, leading to effects at higher levels. Given the current state of knowledge,
exposure indices that cumulate and differentially weight the higher hourly average
concentrations and also include the mid-level values are the most appropriate for use
in developing response functions and comparing studies. The framework for causal
determinations (see Preamble) has been applied to the body of scientific evidence to
examine effects attributed to O3 exposure collectively and the determinations are
presented in Table 9-19.
9-147
-------
Table 9-19 Summary of O3 causal determinations for vegetation and
ecosystem effects.
Vegetation and
Ecosystem Effects
Conclusions from 2006 O3 AQCD
Conclusions from
this ISA
Visible Foliar Injury Effects
on Vegetation
Data published since the 1996 O3 AQCD strengthen previous
conclusions that there is strong evidence that current ambient O3
concentrations cause impaired aesthetic quality of many native plants
and trees by increasing foliar injury.
Causal Relationship
Reduced Vegetation Growth
Data published since the 1996 O3 AQCD strengthen previous
conclusions that there is strong evidence that current ambient O3
concentrations cause decreased growth and biomass accumulation in
annual, perennial and woody plants, including agronomic crops,
annuals, shrubs, grasses, and trees.
Causal Relationship
Reduced Productivity in
Terrestrial Ecosystems
There is evidence that O3 is an important stressor of ecosystems and
that the effects of O3 on individual plants and processes are scaled up
through the ecosystem, affecting net primary productivity.
Causal Relationship
Reduced Carbon (C)
Sequestration in Terrestrial
Ecosystems
Limited studies in the 2006 O3 AQCD.
Likely to be a Causal
Relationship
Reduced Yield and Quality
of Agricultural Crops
Data published since the 1996 O3 AQCD strengthen previous
conclusions that there is strong evidence that current ambient O3
concentrations cause decreased yield and/or nutritive quality in a large
number of agronomic and forage crops.
Causal Relationship
Alteration of Terrestrial
Ecosystem Water Cycling
Ecosystem water quantity may be affected by O3 exposure at the
landscape level.
Likely to be a Causal
Relationship
Alteration of Below-ground
Biogeochemical Cycles
Ozone-sensitive species have well known responses to O3 exposure,
including altered C allocation to below-ground tissues; and also altered
rates of leaf and root production, turnover, and decomposition. These
shifts can affect overall C and N loss from the ecosystem in terms of
respired C, and leached aqueous dissolved organic and inorganic C
and N.
Causal Relationship
Alteration of Terrestrial
Community Composition
Ozone may be affecting above- and below -ground community
composition through impacts on both growth and reproduction.
Significant changes in plant community composition resulting directly
from O3 exposure have been demonstrated.
Likely to be a Causal
Relationship
9-148
-------
References
Agrell. J: Kopper. BJ; McDonald. EP; Lindroth. RL. (2005). CO2 and O3 effects on host plant preferences of
the forest tent caterpillar (Malacosoma disstria). Global Change Biol 11: 588-599.
http://dx.doi.0rg/10.llll/i.1365-2486.2005.00924.x
Ahlfors. R; Brosche. M; Kollist. H; Kangasjarvi. J. (2009). Nitric oxide modulates ozone-induced cell death,
hormone biosynthesis and gene expression in Arabidopsis thaliana. Plant J 58: 1-12.
http://dx.doi.org/10.1111/i. 1365-313X.2008.03756.x
Ahsan. N; Nanjo. Y; Sawada. H; Kohno. Y; Komatsu. S. (2010). Ozone stress-induced proteomic changes in
leaf total soluble and chloroplast proteins of soybean reveal that carbon allocation is involved in adaptation
in the early developmental stage. Proteomics 10: 2605-2619. http://dx.doi.org/10.1002/pmic.201000180
Ainsworth. EA. (2008). Rice production in a changing climate: a meta-analysis of responses to elevated carbon
dioxide and elevated ozone concentration. Global Change Biol 14: 1642-1650.
http://dx.doi.0rg/10.llll/i.1365-2486.2008.01594.x
Ainsworth. EA; Long. SP (2005). What have we learned from 15 years of free-air CO2 enrichment (FACE)?
A meta-analytic review of the responses of photosynthesis, canopy properties and plant production to rising
CO2 [Review]. New Phytol 165: 351-371. http://dx.doi.Org/10.llll/j.1469-8137.2004.01224.x
Ainsworth. EA; Rogers. A. (2007). The response of photosynthesis and stomatal conductance to rising [CO2]:
Mechanisms and environmental interactions [Review]. Plant Cell Environ 30: 258-270.
http://dx.doi.0rg/10.llll/i.1365-3040.2007.01641.x
Alexis. A; Garcia. A; Nystrom. M; Rosenkranz. K. (200la). The 2001 California almanac of emissions and air
quality. Sacremento, CA: California Air Resources Board.
http://www.arb.ca.gov/aqd/almanac/almanac01/almanac01.htm
Allen. EB; Temple. PJ: Bytnerowicz. A; Arbaugh. MJ; Sirulnik. AG: Rao. LE. (2007). Patterns of understory
diversity in mixed coniferous forests of southern California impacted by air pollution.
ScientificWorldJournal 7: 247-263. http://dx.doi.org/10.1100/tsw.2007.72
Alonso. R; Bermejo. V: Sanz. J: Vails. B; Elvira. S: Gimeno. BS. (2007). Stomatal conductance of semi-natural
Mediterranean grasslands: Implications for the development of ozone critical levels. Environ Pollut 146:
692-698. http://dx.doi.0rg/10.1016/i.envpol.2006.06.009
Amthor. JS. (1988). Growth and maintenance respiration in leaves of bean (Phaseolus vulgaris L) exposed to
ozone in open-top chambers in the field. New Phytol 110: 319-325. http://dx.doi.org/10.llll/U469-
8137.1988.tb00268.x
Andersen. CP (2003). Source-sink balance and carbon allocation below ground in plants exposed to ozone.
New Phytol 157: 213-228.
Andersen. CP: Ritter. W: Gregg. J: Matvssek. R; Grams. TEE. (2010). Below-ground carbon allocation in
mature beech and spruce trees following long-term, experimentally enhanced O3 exposure in southern
Germany. Environ Pollut 158: 2604-2609. http://dx.doi.Org/10.1016/i.envpol.2010.05.008
Andersen. CP: Wilson. R; Plocher. M; Hogsett. WE. (1997). Carry-over effects of ozone on root growth and
carbohydrate concentrations of ponderosa pine seedlings. Tree Physiol 17: 805-811.
Andrews. KM; Gibbons. JW: Jochimsen. DM. (2008). Ecological effects of roads on amphibians and reptiles:
A literature review. In JC Mitchell; REJ Brown; B Bartholomew (Eds.), Urban Herpetology (pp. 121-143).
Salt Lake City: Society for the Study of Amphibians and Reptiles.
http://www.uga.edu/srel/Reprint/3091.htm
9-149
-------
Aneja. MK; Sharma. S: Fleischmann. F; Stich. S: Heller. W: Bahnweg. G: Munch. JC: Schloter. M. (2007).
Influence of ozone on litter quality and its subsequent effects on the initial structure of colonizing microbial
communities. Microb Ecol 54: 151-160. http://dx.doi.org/10.1007/s00248-006-9183-0
Arbaugh. M; Bytnerowicz. A; Grulke. N; Fenn. M; Poth. M; Temple. P; Miller. P. (2003). Photochemical smog
effects in mixed conifer forests along a natural gradient of ozone and nitrogen deposition in the San
Bernardino Mountains. Environ Int 29: 401-406. http://dx.doi.org/10.1016/S0160-4120(02)00176-9
Arbaugh. MJ; Miller. PR; Carroll. JJ; Takemoto. BL; Proctor. T. (1998). Relationships of ozone exposure to
pine injury in the Sierra Nevada and San Bernardino Mountains of California, USA. Environ Pollut 101:
291-301. http://dx.doi.org/10.1016/S0269-7491(98)00027-X
Ariyaphanphitak. W: Chidthaisong. A; Sarobol. E; Bashkin. VN; Towpravoon. S. (2005). Effects of elevated
ozone concentrations on Thai Jasmine rice cultivars (Oryza sativa L.). Water Air Soil Pollut 167: 179-200.
http://dx.doi.org/10.1007/sll270-005-8650-4
Ashmore. M; Emberson. L; Karlsson. PE; Pleijel. H. (2004a). Introduction for ozone deposition special issue.
Atmos Environ 38: 2211-2212.
Ashmore. M; Emberson. L; Karlsson. PE; Pleijel. H. (2004b). New directions: A new generation of ozone
critical levels for the protection of vegetation in Europe (correspondence). Atmos Environ 38: 2213-2214.
Ashmore. MR. (2002). Effects of oxidants at the whole plant and community level. In JNB Bell; M Treshow
(Eds.), Air pollution and plant life (pp. 89-118). London: Wiley.
Ashmore. MR; Bell. JNB: Mimmack. A. (1988). Crop growth along a gradient of ambient air pollution.
Environ Pollut 53: 99-121. http://dx.doi.org/10.1016/0269-7491(88)90028-0
Avnery. S: Mauzerall. PL; Liu. J; Horowitz. LW (2011 a). Global crop yield reductions due to surface ozone
exposure: 1. Year 2000 crop production losses and economic damage. Atmos Environ 45: 2284-2296.
http://dx.doi.0rg/10.1016/i.atmosenv.2010.ll.045
Avnery. S: Mauzerall. PL; Liu. J; Horowitz. LW. (20lib). Global crop yield reductions due to surface ozone
exposure: 2. Year 2030 potential crop production losses and economic damage under two scenarios of O3
pollution. Atmos Environ 45: 2297-2309. http://dx.doi.Org/10.1016/i.atmosenv.2011.01.002
Awmack. CS: Harrington. R; Lindroth. RL. (2004). Aphid individual performance may not predict population
responses to elevated CO2 or O3. Global Change Biol Biol.10: 1414-1423.
Awmack. CS: Mondor. EB; Lindroth. RL. (2007). Forest understory clover populations in enriched CO2 and
O-3 atmospheres: Interspecific, intraspecific, and indirect effects. Environ Exp Bot 59: 340-346.
http://dx.doi.0rg/10.1016/i.envexpbot.2006.04.003
Bagard. M; Le Thiec. D; Delacote. E; Hasenfratz-Sauder. MP; Banvoy. J: Gerard. J: Dizengremel. P; Jolivet. Y
(2008). Ozone-induced changes in photosynthesis and photorespiration of hybrid poplar in relation to the
developmental stage of the leaves. Physiol Plant 134: 559-574. http://dx.doi.org/l 0.1111/i. 1399-
3054.2008.01160.x
Baier. M; Kandlbinder. A; Golldack. D; Dietz. K. (2005). Oxidative stress and ozone: Perception; signalling
and response. Plant Cell Environ 28: 1012-1020. http://dx.doi.Org/10.llll/j.1365-3040.2005.01326.x
Baldantoni. D; Fagnano. M; Alfani. A. (2011). Tropospheric ozone effects on chemical composition and
decomposition rate of Quercus ilex L. leaves. Sci Total Environ 409: 979-984.
http://dx.doi.0rg/10.1016/i.scitotenv.2010.ll.022
Ball. GR; Palmer-Brown. D; Fuhrer. J; Skarbv. L; Gimeno. BS; Mills. G. (2000). Identification of non-linear
influences on the seasonal ozone dose-response of sensitive and resistant clover clones using artificial
neural networks. Ecol Modell 129: 153-168.
Balls. GR; Palmer-Brown. D; Sanders. GE. (1996). Investigating microclimatic influences on ozone injury in
clover (Trifolium subterraneum) using artificial neural networks. New Phytol 132: 271-280.
http://dx.doi.0rg/10.llll/i.1469-8137.1996.tb01846.x
9-150
-------
Bandeff. JM; Pregitzer. KS: Loya. WM; Holmes. WE; Zak. DR. (2006). Overstory community composition
and elevated atmospheric CO2 and O3 modify understory biomass production and nitrogen acquisition.
Plant Soil 282: 251-259. http://dx.doi.org/10.1007/slll04-005-5930-0
Barnes. J: Zheng. Y; Lyons. T. (2002). Plant resistance to ozone: The role of ascorbate. In K Omasa; H Saji; S
Youssefian; N Kondo (Eds.), Air pollution and plant biotechnology - Prospects for phytomonitoring and
phytoremediation (pp. 235-252). Tokyo: Springer-Verlag.
Bassin. S: Volk. M; Fuhrer. J. (2007a). Factors affecting the ozone sensitivity of temperate European
grasslands: An overview [Review]. Environ Pollut 146: 678-691.
http://dx.doi.0rg/10.1016/i.envpol.2006.06.010
Bassin. S: Volk. M; Suter. M; Buchmann. N; Fuhrer. J. (2007b). Nitrogen deposition but not ozone affects
productivity and community composition of subalpine grassland after 3 yr of treatment. New Phytol 175:
523-534. http://dx.doi.0rg/10.llll/j.1469-8137.2007.02140.x
Bassin. S: Werner. RA; Sorgel. K; Volk. M; Buchmann. N; Fuhrer. J. (2009). Effects of combined ozone and
nitrogen deposition on the in situ properties of eleven key plant species of a subalpine pasture. Oecologia
158: 747-756. http://dx.doi.org/10.1007/s00442-008-1191-v
Bauer. MR; Hultman. NE; Panek. JA; Goldstein. AH. (2000). Ozone deposition to a ponderosa pine plantation
in the Sierra Nevada Mountains (CA): A comparison of two different climatic years. J Geophys Res 105:
22,123-122,136. http://dx.doi.org/10.1029/2000JD900168
Bender. J: Muntifering. RB; Lin. JC: Weigel. HJ. (2006). Growth and nutritive quality of Poa pratensis as
influenced by ozone and competition. Environ Pollut 142: 109-115.
http://dx.doi.0rg/10.1016/i.envpol.2005.09.012
Bender. J: Weigel. HJ. (2011). Changes in atmospheric chemistry and crop health: A review [Review]. Agron
Sustain Dev 31: 81-89. http://dx.doi.org/10.1051/agro/2010013
Bennett. JP; Jepsen. EA; Roth. JA. (2006). Field responses of Primus serotina and Asclepias syriaca to ozone
around southern Lake Michigan. Environ Pollut 142: 354-366.
http://dx.doi.0rg/10.1016/i.envpol.2005.09.024
Benoit. LF; Skelly. JM; Moore. LD; Dochinger. LS. (1982). Radial growth reductions of Pinus strobus L
correlated with foliar ozone sensitivity as an indicator of ozone-induced losses in eastern forests. Can J For
Res 12: 673-678. http://dx.doi.org/10.1139/x82-101
Bergweiler. CJ: Manning. WJ. (1999). Inhibition of flowering and reproductive success in spreading dogbane
(Apocynum androsaemifolium) by exposure to ambient ozone. Environ Pollut 105: 333-339.
http://dx.doi.org/10.1016/S0269-7491(99)00044-5
Bergweiler. CJ: Manning. WJ: Chevone. BI. (2008). Seasonal and diurnal gas exchange differences in ozone-
sensitive common milkweed (Asclepias syriaca L.) in relation to ozone uptake. Environ Pollut 152: 403-
415. http://dx.doi.0rg/10.1016/j.envpol.2007.06.019
Bernacchi. CJ: Leaky. ADB; Heady. LE; Morgan. PB; Dohleman. FG: McGrath. JM; Gillespie. KM; Wittig.
VE; Rogers. A; Long. SP; Ort. DR. (2006). Hourly and seasonal variation in photosynthesis and stomatal
conductance of soybean grown at future CO2 and ozone concentrations for 3 years under fully open-air
field conditions. Plant Cell Environ 29: 2077-2090. http://dx.doi.Org/10.llll/i.1365-3040.2006.01581.x
Bernacchi. CJ: Morgan. PB; Ort. PR; Long. SP. (2005). The growth of soybean under free air CO2 enrichment
(FACE) stimulates photosynthesis while decreasing in vivo Rubisco capacity. Planta 220: 434-446.
http://dx.doi.org/10.1007/s00425-004-1320-8
Betzelberger. AM; Gillespie. KM; McGrath. JM; Koester. RP; Nelson. RL; Ainsworth. EA. (2010). Effects of
chronic elevated ozone concentration on antioxidant capacity, photosynthesis and seed yield of 10 soybean
cultivars. Plant Cell Environ 33: 1569-1581. http://dx.doi.Org/10.llll/i.1365-3040.2010.02165.x
Bidart-Bouzat. MG: Imeh-Nathaniel. A. (2008). Global change effects on plant chemical defenses against
insect herbivores. J Integr Plant Biol 50: 1339-1354. http://dx.doi.Org/10.llll/i.1744-7909.2008.00751.x
Billings. WD. (1978). Plants and the ecosystem. Belmont, CA: Wadsworth Publishing Company, Inc.
9-151
-------
Biswas. DK; Xu. H; Li. YG: Sun. JZ; Wang. XZ; Han. XG: Jiang. GM. (2008). Genotypic differences in leaf
biochemical, physiological and growth responses to ozone in 20 winter wheat cultivars released over the
past 60 years. Global Change Biol 14: 46-59. http://dx.doi.Org/10.llll/i.1365-2486.2007.01477.x
Black. VJ; Black. CR; Roberts. JA; Stewart. CA. (2000). Impact of ozone on the reproductive development of
plants. New Phytol 147: 421-447.
Black. VJ; Stewart. CA; Roberts. JA; Black. CR. (2007). Ozone affects gas exchange, growth and reproductive
development in Brassica campestris (Wisconsin Fast Plants). New Phytol 176: 150-163.
http://dx.doi.0rg/10.llll/i.1469-8137.2007.02163.x
Black. VJ: Stewart. CA; Roberts. JA: Black. CR. (2010). Direct effects of ozone on reproductive development
in Plantago major L. populations differing in sensitivity. Environ Exp Bot 69: 121-128.
http://dx.doi.0rg/10.1016/i.envexpbot.2010.04.006
Blande. JD; Holopainen. JK; Li. T. (2010). Air pollution impedes plant-to-plant communication by volatiles.
Ecol Lett 13: 1172-1181. http://dx.doi.Org/10.llll/i.1461-0248.2010.01510.x
Border. S: Bagard. M; Oufir. M; Planchon. S: Hoffmann. L; Jolivet Y; Hausman. JF; Dizengremel. P; Renaut.
1 (2007). A DIGE analysis of developing poplar leaves subjected to ozone reveals major changes in carbon
metabolism. Proteomics 7: 1584-1599. http://dx.doi.org/10.1002/pmic.200600822
Bohler. S: Sergeant. K; Lefevre. I; Jolivet. Y; Hoffmann. L; Renaut. J: Dizengremel. P; Hausman. JF. (2010).
Differential impact of chronic ozone exposure on expanding and fully expanded poplar leaves. Tree Physiol
30: 1415-1432. http://dx.doi.org/10.1093/treephys/tpq082
Bonn. B; Von Kuhlmann. R; Lawrence. MG. (2004). High contribution of biogenic hydroperoxides to
secondary organic aerosol formation. Geophys Res Lett 31: L10108.
http://dx.doi.org/10.1029/2003GL019172
Booker. FL; Burkev. KO; Jones. AM. (2012). Re-evaluating the role of ascorbic acid and phenolic glycosides
in ozone scavenging in the leaf apoplast of Arabidopsis thaliana L. Plant Cell Environ 35: 1456-1466.
http://dx.doi.0rg/10.llll/i.1365-3040.2012.02502.x
Booker. FL; Burkev. KO: Overmyer. K; Jones. AM. (2004a). Differential responses of G-protein Arabidopsis
thaliana mutants to ozone. New Phytol 162: 633-641.
Booker. FL; Burkev. KO: Purslev. WA; Heagle. AS. (2007). Elevated carbon dioxide and ozone effects on
peanut: I. Gas-exchange, biomass, and leaf chemistry. Crop Sci 47: 1475-1487.
http://dx.doi.org/10.2135/cropsci2006.08.0537
Booker. FL; Fiscus. EL. (2005). The role of ozone flux and antioxidants in the suppression of ozone injury by
elevated CO2 in soybean. J Exp Bot 56: 2139-2151. http://dx.doi.org/10.1093/ixb/eri214
Booker. FL; Fiscus. EL; Miller. JE. (2004b). Combined effects of elevated atmospheric carbon dioxide and
ozone on soybean whole-plant water use. Environ Manage 33: S355-S362.
http://dx.doi.org/10.1007/s00267-003-9144-z
Booker. FL; Muntifering. R; McGrath. M; Burkev. K; Decoteau. D; Fiscus. E; Manning. W: Krupa. S:
Chappelka. A; Grantz. D. (2009). The ozone component of global change: Potential effects on agricultural
and horticultural plant yield, product quality and interactions with invasive species. J Integr Plant Biol 51:
337-351. http://dx.doi.0rg/10.llll/j.1744-7909.2008.00805.x
Booker. FL; Prior. SA; Torbert. HA; Fiscus. EL; Purslev. WA; Hu. S. (2005). Decomposition of soybean grown
under elevated concentrations of CO2 and O3. Global Change Biol 11: 685-698.
http://dx.doi.0rg/10.llll/i.1365-2486.2005.00939.x
Booker. FL; Reid. CD; Brunschon-Harti. S: Fiscus. EL; Miller. JE. (1997). Photosynthesis and
photorespiration in soybean [Glycine max (L) Merr] chronically exposed to elevated carbon dioxide and
ozone. J Exp Bot 48: 1843-1852.
Borowiak. K; Rucinska-Sobkowiak. R; Rymer. K; Gwozdz. EA; Zbierska. J. (2009). Biochemical markers of
tropospheric ozone: Experimentation with test-plants. Polish Journal of Ecology 57: 3-14.
9-152
-------
Bou Jaoude. M; Katerji. N; Mastrorilli. M; Rana. G. (2008a). Analysis of the effect of ozone on soybean in the
Mediterranean region I: The consequences on crop-water status. Eur J Agron 28: 508-518.
http://dx.doi.0rg/10.1016/i.eja.2007.09.002
Bou Jaoude. M; Katerji. N; Mastrorilli. M; Rana. G. (2008b). Analysis of the ozone effect on soybean in the
Mediterranean region II. The consequences on growth, yield and water use efficiency. Eur J Agron 28: 519-
525. http://dx.doi.0rg/10.1016/i.eia.2007.09.001
Broadmeadow. MSJ: Jackson. SB. (2000). Growth responses of Quercus petraea, Fraxinus excelsior and Pinus
sylvestris to elevated carbon dioxide, ozone and water supply. New Phytol 146: 437-451.
http://dx.doi.0rg/10.1046/i.1469-8137.2000.00665.x
Brook. JR; DiGiovanni. F; Cakmak. S: Meyers. TP. (1997). Estimation of dry deposition velocity using
inferential models and site-specific meteorology-uncertainty due to siting of meteorological towers. Atmos
Environ 31: 3911-3919.
Bulbovas. P; de Souza. SR; de Moraes. PJVI; Luizao. F; Artaxo. P. (2007). Soybean Tracaja' seedlings exposed
to ozone under controlled conditions. Pesqui Agropecu Bras 42: 641-646. http://dx.doi.org/10.1590/S0100-
204X2007000500005
Burkev. KO; Booker. FL; Purslev. WA; Heagle. AS. (2007). Elevated carbon dioxide and ozone effects on
peanut: II. Seed yield and quality. Crop Sci 47: 1488-1497. http://dx.doi.org/10.2135/cropsci2006.08.0538
Burkev. KO: Eason. G: Fiscus. EL. (2003). Factors that affect leaf extracellular ascorbic acid content and
redox status. Physiol Plant 117: 51-57. http://dx.doi.Org/10.1034/i.1399-3054.2003.1170106.x
Burkev. KO: Neufeld. HS: Souza. L; Chappelka. AH; Davison. AW. (2006). Seasonal profiles of leaf ascorbic
acid content and redox state in ozone-sensitive wildflowers. Environ Pollut 143: 427-434.
http://dx.doi.0rg/10.1016/i.envpol.2005.12.009
Bytnerowicz. A; Arbaugh. M; Schilling. S: Fraczek. W: Alexander. D. (2008). Ozone distribution and
phytotoxic potential in mixed conifer forests of the San Bernardino Mountains, Southern California.
Environ Pollut 155: 398-408. http://dx.doi.Org/10.1016/i.envpol.2008.01.046
Caird. MA; Richards. JH; Donovan. LA. (2007). Nighttime stomatal conductance and transpiration in C-3 and
C-4 plants. Plant Physiol 143: 4-10. http://dx.doi.org/10.1104/pp.106.092940
Cal/EPA (California Environmental Protection Agency). (2010). Air quality data branch main page. Available
online at http://www.arb.ca.gov/aqd/aqdpage.htm (accessed January 28, 2011).
Calatavud. A; Alvarado. JW: Barreno. E. (2002). Similar effects of ozone on four cultivars of lettuce in open
top chambers during winter. Photo synthetic a 40: 195-200. http://dx.doi.Org/10.1023/A:1021333305592
Calatavud. A; Pomares. F; Barreno. E. (2006). Interactions between nitrogen fertilization and ozone in
watermelon cultivar Reina de Corazones in open-top chambers. Effects on chlorophyll alpha fluorescence,
lipid peroxidation, and yield. Photosynthetica44: 93-101. http://dx.doi.org/10.1007/sll099-005-0163-2
Calatavud. V: Cervero. J: Sanz. MJ. (2007a). Foliar, physiologial and growth responses of four maple species
exposed to ozone. Water Air Soil Pollut 185: 239-254. http://dx.doi.org/10.1007/sll270-007-9446-5
Calatavud. V: Marco. F; Cervero. J: Sanchez-Peiia. G: Sanz. MJ. (2010). Contrasting ozone sensitivity in
related evergreen and deciduous shrubs. Environ Pollut 158: 3580-3587.
http://dx.doi.0rg/10.1016/i.envpol.2010.08.013
Calatavud. V: Sanz. MJ: Calvo. E; Cervero. J: Ansel. W: Klumpp. A. (2007b). Ozone biomonitoring with Bel-
W3 tobacco plants in the city of Valencia (Spain). Water Air Soil Pollut 183: 283-291.
http://dx.doi.org/10.1007/sll270-007-9376-2
Campbell. SJ: Wanek. R; Coulston. JW. (2007). Ozone injury in west coast forests: 6 years of monitoring -
Introduction. Portland, OR: U.S. Department of Agriculture.
Cannon. WN. (1990). Olfactory response of eastern spruce budworm larvae to red spruce needles exposed to
acid rain and elevated levels of ozone. J Chem Ecol 16: 3255-3261. http://dx.doi.org/10.1007/BF00982096
9-153
-------
Garde. RT; Haynes. KF. (2004). Stucture of the pheromone communication channel in moths. In Advances in
insect chemical ecology. Cambridge: Cambridge University Press.
http://dx.doi.org/10.1017/CBO9780511542664.009
Casteel. CL; O'Neill. BF; Zavala. JA; Bilgin. DP; Berenbaum. MR; DeLucia. EH. (2008). Transcriptional
profiling reveals elevated CO2 and elevated O3 alter resistance of soybean (Glycine max) to Japanese
beetles (Popilliajaponica). Plant Cell Environ 31: 419-434. http://dx.doi.org/10.1111/i. 1365-
3040.2008.01782.x
Chapman. JA; King. JS: Pregitzer. KS: Zak. DR. (2005). Effects of elevated concentrations of atmospheric
CO2 and tropospheric O-3 on decomposition of fine roots. Tree Physiol 25: 1501-1510.
Chappelka. A; Skelly. J: Somers. G: Renfro. J; Hildebrand. E. (1999a). Mature black cherry used as a
bioindicator of ozone injury. Water Air Soil Pollut 116: 261-266.
Chappelka. A; Somers. G: Renfro. J. (1999b). Visible ozone injury on forest trees in Great Smoky Mountains
National Park, USA. Water Air Soil Pollut 116: 255-260.
Chappelka. AH. (2002). Reproductive development of blackberry (Rubus cuneifolius) as influenced by ozone.
New Phytol 155: 249-255. http://dx.doi.Org/10.1046/i.1469-8137.2002.00464.x
Chappelka. AH; Neufeld. HS: Davison. AW: Somers. GL; Renfro. JR. (2003). Ozone injury on cutleaf
coneflower (Rudbeckia laciniata) and crown-beard (Verbesina occidentalis) in Great Smoky Mountains
National Park. Environ Pollut 125: 53-60.
Chappelka. AH; Samuelson. LJ. (1998). Ambient ozone effects on forest trees of the eastern United States:
A review [Review]. New Phytol 139: 91-108. http://dx.doi.Org/10.1046/i.1469-8137.1998.00166.x
Chappelka. AH; Somers. GL; Renfro. JR. (2007). Temporal patterns of foliar ozone symptoms on tall
milkweed (Asclepias exaltata L) in Great Smoky Mountains National Park. Environ Pollut 149: 358-365.
http://dx.doi.0rg/10.1016/i.envpol.2007.05.015
Chen. CW: Tsai. WT; Lucier. AA. (1998). A model of air-tree-soil system for ozone impact analysis. Ecol
Modell 111:207-222.
Chen. Z; Gallie. DR. (2005). Increasing tolerance to ozone by elevating foliar ascorbic acid confers greater
protection against ozone than increasing avoidance. Plant Physiol 138: 1673-1689.
http://dx.doi.org/10.1104/pp.105.062000
Chen. Z; Wang. XK; Feng. ZZ; Xiao. 0: Duan. XN. (2009). Impact of elevated O3 on soil microbial
community function under wheat crop. Water Air Soil Pollut 198: 189-198.
http://dx.doi.org/10.1007/sll270-008-9838-l
Chen. Z; Wang. XK; Yao. FF; Zheng. FX; Feng. ZZ. (201 Ob). Elevated ozone changed soil microbial
community in a rice paddy. Soil Sci Soc Am J 74: 829-837. http://dx.doi.org/10.2136/sssai2009.0258
Cheng. FY; Burkev. KO; Robinson. JM; Booker. FL. (2007). Leaf extracellular ascorbate in relation to O3
tolerance of two soybean cultivars. Environ Pollut 150: 355-362.
http://dx.doi.0rg/10.1016/i.envpol.2007.01.022
Cho. K; Shibato. J: Agrawal. GK; Jung. YH; Kubo. A; Jwa. NS: Tamogami. S: Satoh. K; Kikuchi. S: Higashi.
T; Kimura. S: Saii. H; Tanaka. Y; Iwahashi. H; Masuo. Y; Rakwal. R. (2008). Integrated transcriptomics,
proteomics, and metabolomics analyses to survey ozone responses in the leaves of rice seedling. J
Proteome Res 7: 2980-2998. http://dx.doi.org/10.1021/pr800128q
Christman. MA; Donovan. LA; Richards. JH. (2009). Magnitude of nighttime transpiration does not affect
plant growth or nutrition in well-watered Arabidopsis. Physiol Plant 136: 264-273.
http://dx.doi.0rg/10.llll/i.1399-3054.2009.01216.x
Chung. HG: Zak. PR; Lilleskov. EA. (2006). Fungal community composition and metabolism under elevated
CO2 and O3. Oecologia 147: 143-154. http://dx.doi.org/10.1007/s00442-005-0249-3
Colls. JJ: Unsworth. MH. (1992). Air pollution interactions with natural stressors. In JR Barker; DT Tingey
(Eds.), Air pollution effects on biodiversity. New York, NY: Van Nostrand Reinhold.
9-154
-------
Cornelissen. T. (2011). Climate change and its effects on terrestrial insects and herbivory patterns [Review].
Neotrop Entomol 40: 155-163. http://dx.doi.org/10.1590/S1519-566X2011000200001
Costa. PL; Folinsbee. LJ; Raub. JA; Tilton. B; Tingev. DT. (1992). Summary of selected new information on
effects of ozone on health and vegetation: Supplement to 1986 air quality criteria for ozone and other
photochemical oxidants. (EPA/600/8-88/105F). Research Triangle Park, NC: U.S. Environmental
Protection Agency. http://cfpub.epa.gov/ncea/isa/recordisplav.cfm?deid=31093
Coulston. JW: Smith. GC: Smith. WD. (2003). Regional assessment of ozone sensitive tree species using
bioindicator plants. Environ Monit Assess 83: 113-127.
Crous. KY; Vandermeiren. K; Ceulemans. R. (2006). Physiological responses to cumulative ozone uptake in
two white clover (Trifolium repens L. cv. Regal) clones with different ozone sensitivity. Environ Exp Bot
58: 169-179. http://dx.doi.Org/10.1016/i.envexrjbot.2005.07.007
D'Haese. D; Vandermeiren. K; Asard. H; Horemans. N. (2005). Other factors than apoplastic ascorbate
contribute to the differential ozone tolerance of two clones of Trifolium repens L. Plant Cell Environ 28:
623-632. http://dx.doi.org/10.1111/i. 1365-3040.2005.01308.x
Dalstein. L; Vas. N. (2005). Ozone concentrations and ozone-induced symptoms on coastal and alpine
mediterranean pines in southern France. Water Air Soil Pollut 160: 181-195.
Dammgen. U; Grunhage. L; Haenel. HP; Jager. HJ. (1993). Climate and stress in ecotoxicology: A coherent
system of definitions and terms. J Appl Bot Food Qual 67: 157-162.
Darbah. JNT; Kubiske. ME; Neilson. N; Oksanen. E; Vaapavuori. E; Karnosky. DF (2007). Impacts of
elevated atmospheric CO2 and O3 on paper birch (Betula papyrifera): Reproductive fitness.
ScientificWorldJournal 7: 240-246. http://dx.doi.org/10.1100/tsw.2007.42
Darbah. JNT: Kubiske. ME; Nelson. N; Oksanen. E; Vapaavuori. E; Kamosky. DF. (2008). Effects of decadal
exposure to interacting elevated CO2 and/or O-3 on paper birch (Betula papyrifera) reproduction. Environ
Pollut 155: 446-452. http://dx.doi.Org/10.1016/j.envpol.2008.01.033
Davidson. A. (1993). Update of ozone trends in California's South Coast Air Basin. Air Waste 43: 226-227.
http://dx.doi.org/10.1080/1073161X.1993.10467130
Davis. DP. (2007a). Ozone-induced symptoms on vegetation within the Moosehorn National Wildlife Refuge
in Maine. Northeast Nat 14: 403-414. http://dx.doi.org/10.1656/1092-
6194(2007)14r403:OSOVWT12.0.CO:2
Davis. DP. (2007b). Ozone injury to plants within the Seney National Wildlife Refuge in northern Michigan.
Northeast Nat 14: 415-424.
Davis. DP. (2009). Ozone-induced stipple on plants in the Cape Remain National Wildlife Refuge, South
Carolina. Southeastern Naturalist 8: 471-478.
Pavis. PP; Orendovici. T. (2006). Incidence of ozone symptoms on vegetation within a National Wildlife
Refuge in New Jersey, USA. Environ Pollut 143: 555-564. http://dx.doi.Org/10.1016/i.envpol.2005.10.051
Pawson. TE; Burgess. SS: Tu. KP; Oliveira. RS: Santiago. LS: Fisher. JB; Simonin. KA; Ambrose. AR.
(2007). Nighttime transpiration in woody plants from contrasting ecosystems. Tree Physiol 27: 561-575.
http://dx.doi.0rg/10.1093/treephys/27.4.561
de Lourdes de Bauer. M; Hernandez-Tejeda. T. (2007). A review of ozone-induced effects on the forests of
central Mexico [Review]. Environ Pollut 147: 446-453. http://dx.doi.Org/10.1016/i.envpol.2006.12.020
Pe Temmerman. L; Legrand. G: Vandermeiren. K. (2007). Effects of ozone on sugar beet grown in open-top
chambers. Eur J Agron 26: 1-9. http://dx.doi.Org/10.1016/i.eja.2006.08.001
Pegl'Innocenti. E; Guidi. L; Soldatini. GF. (2007). Effects of elevated ozone on chlorophyll a fluorescence in
symptomatic and asymptomatic leaves of two tomato genotypes. Biol Plantarum 51: 313-321.
http://dx.doi.org/10.1007/sl0535-007-0061-5
9-155
-------
Dermodv. O: O'Neill. BF; Zangerl. AR; Berenbaum. MR; DeLucia. EH. (2008). Effects of elevated CO2 and
O3 on leaf damage and insect abundance in a soybean agroecosystem. Arthropod-Plant Inte 2: 125-135.
Di Baccio. D; Castagna. A; Paoletti. E; Sebastiani. L; Ranieri. A. (2008). Could the differences in O3
sensitivity between two poplar clones be related to a difference in antioxidant defense and secondary
metabolic response to O3 influx? Tree Physiol 28: 1761-1772.
Dickson. RE: Lewin. KF; Isebrands. JG: Coleman. MD; Heilman. WE; Riemenschneider. DE; Sober. J; Host.
GE; Zak. PR; Hendrev. GR; Pregitzer. KS: Karnosky. DF. (2000). Forest Atmosphere Carbon Transfer and
Storage (FACTS-II) the Aspen Free-Air CO2 and O3 Enrichment (FACE) project: An overview. (General
Technical Report NC-214). St. Paul, MN: U.S. Dept. of Agriculture, Forest Service.
http://nrs.fs.fed.us/pubs/278
Ditchkoff. SS: Lewis. JS: Lin. JC: Muntifering. RB; Chappelka. AH. (2009). Nutritive quality of highbush
blackberry (Rubus argutus) exposed to tropospheric ozone. Rangeland Ecol Manag 62: 364-370.
Dizengremel. P; Le Thiec. D; Bagard. M; Jolivet. Y. (2008). Ozone risk assessment for plants: Central role of
metabolism-dependent changes in reducing power. Environ Pollut 156: 11-15.
http://dx.doi.org/10.1016/i.envpol.20Q7.12.024
Dizengremel. P; Le Thiec. D: Hasenfratz-Sauder. MP; Vaultier. MN; Bagard. M; Jolivet. Y. (2009). Metabolic-
dependent changes in plant cell redox power after ozone exposure. Plant Biol (Stuttg) 11: 35-42.
http://dx.doi.0rg/10.llll/i.1438-8677.2009.00261.x
Dizengremel. P: Sasek. T; Brown. K; Richardson. C. (1994). Ozone-induced changes in primary carbon
metabolism enzymes of loblolly pine needles. J Plant Physiol 144: 300-306.
Dobson. HEM. (1994). Floral volatiles in insect biology. In EA Bernays (Ed.), Insect-plant interactions: Vol 5
(pp. 47-82). Boca Raton, FL: CRC Press.
Dohm. MR; Mautz. WJ: Andrade. JA; Gellert. KS: Salas-Ferguson. LJ; Nicolaisen. N; Fujie. N. (2005). Effects
of ozone exposure on nonspecific phagocytic capacity of pulmonary macrophages from an amphibian,
Bufo marinus. Environ Toxicol Chem 24: 205-210.
Dohm. MR; Mautz. WJ: Doratt. RE; Stevens. JR. (2008). Ozone exposure affects feeding and locomotor
behavior of adult Bufo marinus. Environ Toxicol Chem 27: 1209-1216. http://dx.doi.Org/10.1897/07-388.l
Dohm. MR; Mautz. WJ: Looby. PG: Gellert. KS: Andrade. JA. (2001). Effects of ozone on evaporative water
loss and thermoregulatory behavior of marine toads (Bufo marinus). Environ Res 86: 274-286.
Dohrmann. AB; Tebbe. CC. (2005). Effect of elevated tropospheric ozone on the structure of bacterial
communities inhabiting the rhizosphere of herbaceous plants native to Germany. Appl Environ Microbiol
71: 7750-7758. http://dx.doi.org/10.1128/AEM.71.12.7750-7758.2005
Drogoudi. PD; Ashmore. M. (2001). 14C-allocation of flowering and deblossomed strawberry in response to
elevated ozone. New Phytol 152: 455-461. http://dx.doi.Org/10.1046/i.0028-646X.2001.00270.x
Drogoudi. PD: Ashmore. MR. (2000). Does elevated ozone have differing effects in flowering and
deblossomed strawberry? New Phytol 147: 561-569. http://dx.doi.Org/10.1046/i.1469-8137.2000.00718.x
Dudareva. N; Negre. F; Nagegowda. DA; Orlova. I. (2006). Plant volatiles: Recent advances and future
perspectives [Review]. Crit Rev Plant Sci 25: 417-440.
Ederli. L; Morettini. R; Borgogni. A; Wasternack. C: Miersch. O; Reale. L; Ferranti. F; Tosti. N; Pasqualini. S.
(2006). Interaction between nitric oxide and ethylene in the induction of alternative oxidase in ozone-
treated tobacco plants. Plant Physiol 142: 595-608. http://dx.doi.org/10.1104/pp.106.085472
Edwards. IP; Zak. DR. (2011). Fungal community composition and function after long-term exposure of
northern forests to elevated atmospheric CO2 and tropospheric O3. Global Change Biol 17: 2184-2195.
http://dx.doi.0rg/10.llll/i.1365-2486.2010.02376.x
Ellenson. JL; Amundson. RG. (1982). Delayed light imaging for the early detection of plant stress. Science
215: 1104-1106. http://dx.doi.org/10.1126/science.215.4536.1104
9-156
-------
Ellsworth. PS: Reich. PB; Naumburg. ES: Koch. GW: Kubiske. ME; Smith. SD. (2004). Photosynthesis,
carboxylation and leaf nitrogen responses of 16 species to elevated pCO2 across four free-air CO2
enrichment experiments in forest, grassland and desert. Global Change Biol 10: 2121-2138.
http://dx.doi.0rg/10.llll/i.1365-2486.2004.00867.x
Eltaveb. AE; Kawano. N; Badawi. GH; Kaminaka. H; Sanekata. T; Morishima. I; Shibahara. T; Inanaga. S:
Tanaka. K. (2006). Enhanced tolerance to ozone and drought stresses in transgenic tobacco overexpressing
dehydroascorbate reductase in cytosol. Physiol Plant 127: 57-65. http://dx.doi.org/10.1111/i. 1399-
3054.2005.00624.x
Eltaveb. AE; Kawano. N; Badawi. GH; Kaminaka. H; Sanekata. T; Shibahara. T; Inanaga. S: Tanaka. K.
(2007). Overexpression of monodehydroascorbate reductase in transgenic tobacco confers enhanced
tolerance to ozone, salt and polyethylene glycol stresses. Planta 225: 1255-1264.
http://dx.doi.org/10.1007/s00425-006-0417-7
Emberson. L; Ashmore. MR; Cambridge. HM; Simpson. D; Tuovinen. JP (2000a). Modelling stomatal ozone
flux across Europe. Environ Pollut 109: 403-413. http://dx.doi.org/10.1016/80269-7491(00)00043-9
Emberson. LD; Wieser. G: Ashmore. MR. (2000b). Modelling of stomatal conductance and ozone flux of
Norway spruce: Comparison with field data. Environ Pollut 109: 393-402. http://dx.doi.org/10.1016/S0269-
7491(00)00042-7
Enders. G. (1992). Deposition of ozone to a mature spruce forest: Measurements and comparison to models.
Environ Pollut 75: 61-67. http://dx.doi.org/10.1016/0269-7491(92)90057-H
Esperschutz. J; Pritsch. K; Gattinger. A; Welzl. G: Haesler. F; Buegger. F; Winkler. JB; Munch. JC: Schloter.
M. (2009). Influence of chronic ozone stress on carbon translocation pattern into rhizosphere microbial
communities of beech trees (Fagus sylvatica L.) during a growing season. Plant Soil 323: 85-95.
http://dx.doi.org/10.1007/slll04-009-0090-2
Fares. S: Barta. C: Brilli. F; Centritto. M; Ederli. L; Ferranti. F; Pasqualini. S: Reale. L; Tricoli. D; Loreto. F.
(2006). Impact of high ozone on isoprene emission, photosynthesis and histology of developing Populus
alba leaves directly or indirectly exposed to the pollutant. Physiol Plant 128: 456-465.
http://dx.doi.0rg/10.llll/i.1399-3054.2006.00750.x
Fares. S: Gentner. PR; Park. JH; Ormeno. E; Karlik. J; Goldstein. AH. (2011). Biogenic emissions from Citrus
species in California. Atmos Environ 45: 4557-4568. http://dx.doi.org/10.1016/j.atmosenv.2011.05.066
Fares. S: Loreto. F; Kleist. E; Wildt. J. (2008). Stomatal uptake and stomatal deposition of ozone in isoprene
and monoterpene emitting plants. Plant Biol (Stuttg) 10: 44-54. http://dx.doi.org/10.1055/s-2007-965257
Fares. S: McKay. M; Holzinger. R; Goldstein. AH. (2010a). Ozone fluxes in a Pinus ponderosa ecosystem are
dominated by non-stomatal processes: Evidence from long-term continuous measurements. Agr Forest
Meteorol 150: 420-431. http://dx.doi.Org/10.1016/j.agrformet.2010.01.007
Fares. S: Oksanen. E; Lannenpaa. M; Julkunen-Tiitto. R; Loreto. F. (201 Ob). Volatile emissions and phenolic
compound concentrations along a vertical profile of Populus nigra leaves exposed to realistic ozone
concentrations. Photosynth Res 104: 61-74. http://dx.doi.org/10.1007/slll20-010-9549-5
Felzer. B; Kicklighter. D; Melillo. J: Wang. C: Xhuang. Q; Prinn. R. (2004). Effects of ozone on net primary
production and carbon sequestration in the conterminous United States using a biogeochemistry model.
Tellus B Chem Phys Meteorol 56: 230-248. http://dx.doi.Org/10.llll/i.1600-0889.2004.00097.x
Felzer. B; Reilly. J: Melillo. J: Kicklighter. D; Sarofim. M; Wang. C: Prinn. R; Zhuang. Q. (2005). Future
effects of ozone on carbon sequestration and climate change policy using a global biogeochemical model.
Clim Change 73: 345-373. http://dx.doi.org/10.1007/sl0584-005-6776-4
Felzer. BS: Cronin. TW: Melillo. JM; Kicklighter. DW: Schlosser. CA. (2009). Importance of carbon-nitrogen
interactions and ozone on ecosystem hydrology during the 21st century. J Geophys Res 114: G01020.
Feng. YW: Komatsu. S: Furukawa. T; Koshiba. T; Kohno. Y. (2008a). Proteome analysis of proteins
responsive to ambient and elevated ozone in rice seedlings. Agric Ecosyst Environ 125: 255-265.
http://dx.doi.0rg/10.1016/i.agee.2008.01.018
9-157
-------
Feng. Z; Pang. J; Nouchi. I; Kobavashi. K; Yamakawa. T; Zhu. J. (2010). Apoplastic ascorbate contributes to
the differential ozone sensitivity in two varieties of winter wheat under fully open-air field conditions.
Environ Pollut 158: 3539-3545. http://dx.doi.Org/10.1016/j.envpol.2010.08.019
Feng. ZZ; Kobavashi. K. (2009). Assessing the impacts of current and future concentrations of surface ozone
on crop yield with meta-analysis. Atmos Environ 43: 1510-1519.
http://dx.doi.0rg/10.1016/i.atmosenv.2008.ll.033
Feng. ZZ; Kobavashi. K; Ainsworth. EA. (2008b). Impact of elevated ozone concentration on growth,
physiology, and yield of wheat (Triticum aestivum L.): A meta-analysis. Global Change Biol 14: 2696-
2708. http://dx.doi.0rg/10.llll/i.1365-2486.2008.01673.x
Fenn. ME; de Bauer. LI; Hernandez-Tejeda. T. (2002). Summary of air pollution impacts on forests in the
Mexico City air basin. In Urban air pollution and forests. New York, NY: Springer-Verlag.
Fenn. ME; Poth. MA; Johnson. DW. (1996). Evidence for nitrogen saturation in the San Bernardino Mountains
in southern California. For Ecol Manage 82: 211-230. http://dx.doi.org/10.1016/0378-1127(95)03668-7
Findlev. DA; Keever. GJ: Chappelka. AH; Bakes. DJ: Gillian. DJ. (1997). Differential responses of buddleia
(Buddleia davidii Franch) to ozone. Environ Pollut 98: 105-111.
Finkelstein. PL; Ellestad. TG: Clarke. JF; Meyers. TP; Schwede. DB; Hebert. EO; Neal. JA. (2000). Ozone
and sulfur dioxide dry deposition to forests: Observations and model evaluation. J Geophys Res 105:
15365-15377.
Finnan. JM; Burke. JL; Jones. MB. (1997). An evaluation of indices that describe the impact of ozone on the
yield of spring wheat (Triticum aestivum L). Atmos Environ 31: 2685-2693.
http://dx.doi.org/10.1016/81352-2310(97)00105-2
Finnan. JM; Jones. MB; Burke. JL (1996). A time-concentration study on the effects of ozone on spring wheat
(Triticum aestivum L): 2. A comparison of indices. Agric Ecosyst Environ 57: 169-177.
http://dx.doi.org/10.1016/0167-8809(95)01004-l
Fiscus. EL; Booker. FL; Burkev. KO. (2005). Crop responses to ozone: Uptake, modes of action, carbon
assimilation and partitioning. Plant Cell Environ 28: 997-1011.
Fiscus. EL; Philbeck. R; Britt. AM; Booker. FL. (1999). Growth of Arabidopsis flavonoid mutants under solar
radiation and UV filters. Environ Exp Bot 41: 231-245. http://dx.doi.org/10.1016/80098-8472(99)00011-8
Fishman. J: Bowman. KW: Burrows. JP; Richter. A; Chance. KV; Edwards. DP; Martin. RV: Morris. GA;
Pierce. RB; Ziemke. JR; Al-Saadi. JA; Creilson. JK; Schaack. TK; Thompson. AM. (2008). Remote sensing
of tropospheric pollution from space. Bull Am Meteorol Soc 89: 805-821.
http://dx.doi.Org/10.1175/2008BAMS2526.l
Fishman. J: Creilson. JK; Parker. PA; Ainsworth. EA; Vining. GG: Szarka. J: Booker. FL; Xu. XJ. (2010).
An investigation of widespread ozone damage to the soybean crop in the upper Midwest determined from
ground-based and satellite measurements. Atmos Environ 44: 2248-2256.
http://dx.doi.0rg/10.1016/i.atmosenv.2010.01.015
Flagler. RB. (1998). Recognition of air pollution injury to vegetation: A pictorial atlas (2nd ed.). Pittsburgh,
PA: Air & Waste Management Association.
Flowers. MD; Fiscus. EL; Burkev. KO: Booker. FL; Dubois. JJB. (2007). Photosynthesis, chlorophyll
fluorescence, and yield of snap bean (Phaseolus vulgaris L.) genotypes differing in sensitivity to ozone.
Environ Exp Bot 61: 190-198. http://dx.doi.Org/10.1016/i.envexpbot.2007.05.009
Fontan. JA: Minga. A; Lopez. A; Druilhet A. (1992). Vertical ozone profiles in a pine forest. Atmos Environ
26: 863-869. http://dx.doi.org/10.1016/0960-1686(92)90245-G
Foyer. CH; Noctor. G. (2005 a). Oxidant and antioxidant signalling in plants: Are-evaluation of the concept of
oxidative stress in a physiological context. Plant Cell Environ 28: 1056-1071.
9-158
-------
Foyer. CH; Noctor. G. (2005b). Redox homeostasis and antioxidant signaling: A metabolic interface between
stress perception and physiological responses [Review]. Plant Cell 17: 1866-1875.
http://dx.doi.org/10.1105/tpc.105.033589
Fredericksen. TS: Joyce. BJ; Skelly. JM; Steiner. KG: Kolb. TE; Kouterick. KB; Savage. JE; Snyder. KR.
(1995). Physiology, morphology, and ozone uptake of leaves of black cherry seedlings, saplings, and
canopy trees. Environ Pollut 89: 273-283. http://dx.doi.org/10.1016/0269-7491(94)00077-Q
Fredericksen. TS: Kolb. TE; Skelly. JM; Steiner. KG: Joyce. BJ: Savage. JE. (1996). Light environment alters
ozone uptake per net photosynthetic rate in black cherry trees. Tree Physiol 16: 485-490.
Freiwald. V: Haikio. E; Julkunen-Tiitto. R; Holopainen. JK; Oksanen. E. (2008). Elevated ozone modifies the
feeding behaviour of the common leaf weevil on hybrid aspen through shifts in developmental, chemical,
and structural properties of leaves. Entomol Exp Appl 128: 66-72. http://dx.doi.org/l0.1111/i .1570-
7458.2008.00677.x
Fuentes. JD; Gillespie. TJ; den Hartog. G: Neumann. HH. (1992). Ozone deposition onto a deciduous forest
during dry and wet conditions. Agr Forest Meteorol 62: 1-18. http://dx.doi.org/10.1016/0168-
1923(92)90002-L
Fuhrer. J. (1994). Effects of ozone on managed pasture: 1. Effects of open-top chambers on microclimate,
ozone flux, and plant growth. Environ Pollut 86: 297-305.
Fuhrer. J; Skarbv. L; Ashmore. MR. (1997). Critical levels for ozone effects on vegetation in Europe. Environ
Pollut 97: 91-106. http://dx.doi.org/10.1016/S0269-7491(97)00067-5
Gate. IM; McNeill. S: Ashmore. MR. (1995). Effects of air pollution on the searching behaviour of an insect
parasitoid. Water Air Soil Pollut 85: 1425-1430. http://dx.doi.org/10.1007/BF00477181
Geiser. LH; Neitlich. PN. (2007). Air pollution and climate gradients in western Oregon and Washington
indicated by epiphytic macrolichens. Environ Pollut 145: 203-218.
http://dx.doi.0rg/10.1016/i.envpol.2006.03.024
Gerosa. G: Marzuoli. R; Rossini. M; Panigada. C: Meroni. M; Colombo. R; Faoro. F; Iriti. M. (2009). A flux-
based assessment of the effects of ozone on foliar injury, photosynthesis, and yield of bean (Phaseolus
vulgaris L. cv. Borlotto Nano Lingua di Fuoco) in open-top chambers. Environ Pollut 157: 1727-1736.
http://dx.doi.0rg/10.1016/i.envpol.2008.06.028
Gielen. B; Vandermeiren. K; Horemans. N; D'Haese. D; Serneels. R: Valcke. R. (2006). Chlorophyll a
fluorescence imaging of ozone-stressed Brassica napus L. plants differing in glucosinolate concentrations.
Plant Biol (Stuttg) 8: 698-705. http://dx.doi.org/10.1055/s-2006-924150
Gitav. H; Brown. S: Easterling. W: Jallow. B. (2001). Ecosystems and their goods and services. In Climate
change 2001: Impacts, adaptation and vulnerability: Contribution of Working Group II to the third
assessment report of the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom:
Cambridge University Press.
Gombert. S: Asta. J: Seaward. MRP. (2006). Lichens and tobacco plants as complementary biomonitors of air
pollution in the Grenoble area (Isere, southeast France). Ecol Indicat 6: 429-443.
http://dx.doi.0rg/10.1016/i.ecolind.2005.06.001
Gonzalez-Fernandez. I; Kaminska. A; Dodmani. M; Goumenaki. E; Quarrie. S: Barnes. JD. (2010).
Establishing ozone flux-response relationships for winter wheat: Analysis of uncertainties based on data for
UK and Polish genotypes. Atmos Environ 44: 621-630. http://dx.doi.Org/10.1016/i.atmosenv.2009.ll.021
Goumenaki. E; Tavbi. T; Borland. A; Barnes. J. (2010). Mechanisms underlying the impacts of ozone on
photosynthetic performance. Environ Exp Bot 69: 259-266.
http://dx.doi.0rg/10.1016/i.envexpbot.2010.04.011
Grantz. DA; Gunn. S: Vu. HB. (2006). O3 impacts on plant development: A meta-analysis of root/shoot
allocation and growth. Plant Cell Environ 29: 1193-1209. http://dx.doi.org/10.1111/i. 1365-
3040.2006.01521.x
9-159
-------
Grantz. DA; Shrestfaa. A. (2006). Tropospheric ozone and interspecific competition between yellow nutsedge
and pima cotton. Crop Sci 46: 1879-1889. http://dx.doi.org/10.2135/cropsci2005.06.0167
Grantz. DA; Shrestha. A; Vu. HB. (2008). Early vigor and ozone response in horseweed (Conyza canadensis)
biotypes differing in glyphosate resistance. Weed Sci 56: 224-230. http://dx.doi.Org/10.1614/ws-07-130.l
Grantz. DA; Shrestha. A; Vu. HB. (2010a). Ozone impacts on assimilation and allocation to reproductive sinks
in the vegetatively propagated C-4 weed, yellow nutsedge. Crop Sci 50: 246-252.
http://dx.doi.org/10.2135/cropsci2009.03.0127
Grantz. DA; Vu. HB. (2009). O3 sensitivity in a potential C4 bioenergy crop: Sugarcane in California. Crop
Sci 49: 643-650.
Grantz. DA; Vu. HB: Aguilar. C: Rea. MA. (2010b). No interaction between methyl jasmonate and ozone in
Pima cotton: Growth and allocation respond independently to both. Plant Cell Environ 33: 717-728.
http://dx.doi.0rg/10.llll/i.1365-3040.2009.02096.x
Grantz. DA; Zhang. XJ; Massman. W: Delany. A; Pederson. R. (1997). Ozone deposition to a cotton
(Gossypium hirsutum L) field: Stomatal and surface wetness effects during the California Ozone
Deposition experiment. Agr Forest Meteorol 85: 19-31. http://dx.doi.org/10.1016/80168-1923(96)02396-9
Grantz. DA; Zhang. XJ: Massman. WJ: Den Hartog. G: Neumann. HH; Pederson. JR. (1995). Effects of
stomatal conductance and surface wetness on ozone deposition in field-grown grape. Atmos Environ 29:
3189-3198. http://dx.doi.org/10.1016/1352-2310(95)00129-M
Grebenc. T; Kraigher. H. (2007). Changes in the community of ectomycorrhizal fungi and increased fine root
number under adult beech trees chronically fumigated with double ambient ozone concentration. Plant Biol
(Stuttg) 9: 279-287. http://dx.doi.org/10.1055/s-2006-924489
Gregg. JW: Jones. CG: Dawson. TE. (2003). Urbanization effects on tree growth in the vicinity of New York
City [Letter]. Nature 424: 183-187. http://dx.doi.org/l0.1038/natureO 1728
Gregg. JW: Jones. CG: Dawson. TE. (2006). Physiological and developmental effects of O3 on cottonwood
growth in urban and rural sites. Ecol Appl 16: 2368-2381. http://dx.doi.org/10.1890/1051-
0761(2006)016r2368:PADEOO12.0.CO:2
Grennfelt. P. (2004). New directions: Recent research findings may change ozone control policies. Atmos
Environ 38: 2215-2216.
Groppa. MD; Benavides. MP (2008). Polyamines and abiotic stress: Recent advances. Amino Acids 34: 35-45.
http://dx.doi.org/10.1007/s00726-007-0501-8
Grulke. N; Neufeld. H; Davison. A; Roberts. M; Chappelka. A. (2007a). Stomatal behavior of ozone-sensitive
and -insensitive coneflowers (Rudbeckia laciniata var. digitata) in Great Smoky Mountains National Park.
NewPhytol 173: 100-109. http://dx.doi.Org/10.llll/i.1469-8137.2006.01872.x
Grulke. NE. (1999). Physiological responses of ponderosa pine to gradients of environmental stressors. In PR
Miller; JR McBride (Eds.), Oxidant air pollution impacts in the montane forests of Southern California (pp.
126-163). New York, NY: Springer-Verlag.
Grulke. NE; Alonso. R; Nguyen. T; Cascio. C: Dobrowolski. W. (2004). Stomata open at night in pole-sized
and mature ponderosa pine: Implications for O3 exposure metrics. Tree Physiol 24: 1001-1010.
Grulke. NE; Dobrowolski. W: Mingus. P; Fenn. ME. (2005). California black oak response to nitrogen
amendment at a high O3, nitrogen-saturated site. Environ Pollut 137: 536-545.
http://dx.doi.0rg/10.1016/i.envpol.2005.01.039
Grulke. NE; Johnson. R; Esperanza. A; Jones. D; Nguyen. T; Posch. S: Tausz. M. (2003a). Canopy
transpiration of Jeffrey pine in mesic and xeric microsites: O3 uptake and injury response. Trees Struct
Funct 17: 292-298.
Grulke. NE; Johnson. R; Monschein. S: Nikolova. P; Tausz. M. (2003b). Variation in morphological and
biochemical O3 injury attributes of mature Jeffrey pine within canopies and between microsites. Tree
Physiol 23: 923-929.
9-160
-------
Grulke. NE; Lee. EH. (1997). Assessing visible ozone-induced injury in ponderosa pine. Can J For Res 27:
1658-1668.
Grulke. NE; Minnich. RA; Paine. TD; Sevbold. SJ: Chavez. DJ; Fenn. ME; Riggan. PJ; Dunn. A. (2008). Air
pollution increases forest susceptibility to wildfires: A case study in the San Bernardino Mountains in
southern California. In A Bytnerowicz; MJ Arbaugh; AR Riebau; C Anderson (Eds.), Wildland fires and air
pollution; Section III: Ecological Impacts of Forest Fires and Air Pollution (pp. 365-403). Amsterdam,
The Netherlands: Elsevier Ltd. http://dx.doi.org/10.1016/S1474-8177(08)00017-X
Grulke. NE; Paoletti. E; Heath. RL. (2007b). Chronic vs. short-term acute O3 exposure effects on nocturnal
transpiration in two Californian oaks. ScientificWorldJournal 7: 134-140.
http://dx.doi.org/10.1100/tsw.2007.33
Grulke. NE; Paoletti. E; Heath. RL. (2007c). Comparison of calculated and measured foliar O3 flux in crop
and forest species. Environ Pollut 146: 640-647. http://dx.doi.Org/10.1016/i.envpol.2006.04.014
Grulke. NE; Preisler. HK; Rose. C: Kirsch. J; Balduman. L. (2002). O3 uptake and drought stress effects on
carbon acquisition of ponderosa pine in natural stands. New Phytol 154: 621-631.
http://dx.doi.0rg/10.1046/i.1469-8137.2002.00403.x
Grunhage. L; Haenel. HP. (1997). PLATIN (PLant-ATmosphere-INteraction) I: A model of plant-atmosphere
interaction for estimating absorbed doses of gaseous air pollutants. Environ Pollut 98: 37-50.
http://dx.doi.org/10.1016/S0269-7491(97)00114-0
Grunhage. L; Jager. HJ. (2003). From critical levels to critical loads for ozone: a discussion of a new
experimental and modelling approach for establishing flux-response relationships for agricultural crops and
native plant species. Environ Pollut 125: 99-110.
Grunhage. L; Krupa. SV: Legge. AH; Jager. HJ. (2004). Ambient flux-based critical values of ozone for
protecting vegetation: Differing spatial scales and uncertainties in risk assessment. Atmos Environ 38:
2433-2437. http://dx.doi.Org/10.1016/i.atmosenv.2003.12.039
Guenther. A; Karl. T; Harlev. P; Wiedinmyer. C: Palmer. PI; Geron. C. (2006). Estimates of global terrestrial
isoprene emissions using MEGAN (Model of Emissions of Gases and Aerosols from Nature). Atmos Chem
Phys 6: 3181-3210. http://dx.doi.org/10.5194/acp-6-3181-2006
Guidi. L; Degl'Innocenti. E; Martinelli. F; Piras. M. (2009). Ozone effects on carbon metabolism in sensitive
and insensitive Phaseolus cultivars. Environ Exp Bot 66: 117-125.
http://dx.doi.0rg/10.1016/i.envexpbot.2008.12.005
Guidi. L; Degl'Innocenti. E. (2008). Ozone effects on high light-induced photoinhibition in Phaseolus vulgaris.
Plant Sci 174: 590-596. http://dx.doi.Org/10.1016/i.plantsci.2008.03.003
Gill. H; Gaga. EO; Dogeroglu. T; Ozden. O; Ayvaz. O; Ozel. S: Giingor. G. (2011). Respiratory health
symptoms among students exposed to different levels of air pollution in a Turkish city. Int J Environ Res
Public Health 8: 1110-1125. http://dx.doi.org/10.3390/ijerph8041110
Gumpertz. ML; Rawlings. JO. (1992). Nonlinear regression with variance components: Modeling effects of
ozone on crop yield. Crop Sci 32: 219-224.
http://dx.doi.org/10.2135/cropscil992.0011183X003200010045x
Gunderson. CA; Sholtis. JD; Wullschleger. SD; Tissue. DT; Hanson. PJ: Norbv. RJ. (2002). Environmental and
stomatal control of photosynthetic enhancement in the canopy of a sweetgum (Liquidambar styraciflua L)
plantation during 3 years of CO2 enrichment. Plant Cell Environ 25: 379-393.
http://dx.doi.0rg/10.1046/i.0016-8025.2001.00816.x
Haberer. K; Herbinger. K; Alexou. M; Rennenberg. H; Tausz. M. (2008). Effects of drought and canopy ozone
exposure on antioxidants in fine roots of mature European beech (Fagus sylvatica). Tree Physiol 28: 713-
719. http://dx.doi.0rg/10.1093/treephvs/28.5.713
Hamel. LP; Miles. GP; Samuel. MA; Ellis. BE; Seguin. A; Beaudoin. N. (2005). Activation of stress-
responsive mitogen-activated protein kinase pathways in hybrid poplar (Populus trichocarpa x Populus
deltoides). Tree Physiol 25: 277-288. http://dx.doi.Org/10.1093/treephys/25.3.277
9-161
-------
Hamilton. JG: Dermodv. O: Aldea. M; Zangerl. AR; Rogers. A; Berenbaum. MR; DeLucia. EH. (2005).
Anthropogenic changes in tropospheric composition increase susceptibility of soybean to insect herbivory.
Environ Entomol 34: 479-485.
Handlev. T; Grulke. NE. (2008). Interactive effects of O3 exposure on California black oak (Quercus kelloggii
Newb.) seedlings with and without N amendment. Environ Pollut 156: 53-60.
http://dx.doi.0rg/10.1016/i.envpol.2008.01.002
Hanson. PJ; Wullschleger. SD; Norbv. RJ; Tschaplinski. TJ; Gunderson. CA. (2005). Importance of changing
CO2, temperature, precipitation, and ozone on carbon and water cycles of an upland-oak forest:
incorporating experimental results into model simulations. Global Change Biol 11: 1402-1423.
http://dx.doi.0rg/10.llll/i.1365-2486.2005.00991.x
Harward. M; Treshow. M. (1975). Impact of ozone on the growth and reproduction of understorey plants in the
Aspen zone of western USA. Environ Conserv 2: 17-23. http://dx.doi.org/10.1017/S0376892900000564
Hassan. R; Scholes. R; Ash. N. (2005). Ecosystems and human well-being: Current state and trends.
Washington, DC: Island Press.
Haves. F; Jones. MLM; Mills. G: Ashmore. M. (2007). Meta-analysis of the relative sensitivity of semi-natural
vegetation species to ozone. Environ Pollut 146: 754-762. http://dx.doi.Org/10.1016/j.envpol.2006.06.011
Haves. F; Mills. G: Ashmore. M. (2009). Effects of ozone on inter- and intra-species competition and
photosynthesis in mesocosms of Lolium perenne and Trifolium repens. Environ Pollut 157: 208-214.
http://dx.doi.0rg/10.1016/i.envpol.2008.07.002
Haves. F; Mills. G: Ashmore. M. (2010). How Much Does the Presence of a Competitor Modify the Within-
Canopy Distribution of Ozone-Induced Senescence and Visible Injury? Water Air Soil Pollut 210: 265-276.
http://dx.doi.org/10.1007/sll270-009-0248-9
He. XY; Fu. SL; Chen. W: Zhao. TH; Xu. S: Tuba. Z. (2007). Changes in effects of ozone exposure on growth,
photosynthesis, and respiration of Ginkgo biloba in Shenyang urban area. Photosynthetica 45: 555-561.
http://dx.doi.org/10.1007/sll099-007-0095-0
He. XY; Ruan. YN; Chen. W: Lu. T. (2006). Responses of anti-oxidative system in leaves of Ginkgo biloba to
elevated ozone concentration in urban area. Botanical Studies 47: 409-416.
Heagle. AS. (1979). Effects of growth media, fertiliser rate and hour and season of exposure on sensitivity of
four soybean cultivars to ozone. Environ Pollut 18: 313-322.
Heagle. AS. (1989). Ozone and crop yield*. Annu Rev Phytopathol 27: 397-423.
http://dx.doi.org/10.1146/annurev.pv.27.090189.002145
Heagle. AS: Body. DE; Heck. WW. (1973). An open-top field chamber to assess the impact of air pollution on
plants. J Environ Qua! 2: 365-368.
Heagle. AS: Brandenburg. RL; Burns. JC: Miller. JE. (1994a). Ozone and carbon dioxide effects on spider
mites in white clover and peanut. J Environ Qua! 23: 1168-1176.
http://dx.doi.org/10.2134/ieql994.00472425002300060006x
Heagle. AS: Heck. WW: Lesser. VM; Rawlings. JO. (1987). Effects of daily ozone exposure duration and
concentration fluctuation on yield of tobacco. Phytopathology 77: 856-862.
http://dx.doi.org/10.1094/Phvto-77-856
Heagle. AS: Kress. LW: Temple. PJ: Kohut RJ: Miller. JE; Heggestad. HE. (1988). Factors influencing ozone
dose-yield response relationships in open-top field chamber studies. In WW Heck; OC Taylor; DT Tingey
(Eds.), Assessment of crop loss from air pollutants: Proceedings of an international conference (pp. 141-
179). New York, NY: Elsevier Applied Science.
Heagle. AS; Letchworth. MB; Mitchell. CA. (1983). Effects of growth medium and fertilizer rate on the yield
response of soybeans exposed to chronic doses of ozone. Phytopathology 73: 134-139.
http://dx.doi.org/10.1094/Phvto-73-134
9-162
-------
Heagle. AS: Miller. JE; Rawlings. JO: Vozzo. SF. (1991). Effect of growth stage on soybean response to
chronic ozone exposure. J Environ Qual 20: 562-570.
http://dx.doi.org/10.2134/ieql991.00472425002000030010x
Heagle. AS: Miller. JE; Sherrill. DE. (1994b). A white clover system to estimate effects of tropospheric ozone
on plants. J Environ Qual 23: 613-621. http://dx.doi.org/10.2134/ieal994.00472425002300030030x
Heagle. AS: Reinert. RA; Miller. JE. (1996). Response of white clover to ozone in different environments. J
Environ Qual 25: 273-278. http://dx.doi.org/10.2134/ieql996.00472425002500020010x
Heath. RL. (2008). Modification of the biochemical pathways of plants induced by ozone: What are the varied
routes to change? Environ Pollut 155: 453-463. http://dx.doi.Org/10.1016/i.envpol.2008.03.010
Heath. RL; Lefohn. AS: Musselman. RC. (2009). Temporal processes that contribute to nonlinearity in
vegetation responses to ozone exposure and dose. Atmos Environ 43: 2919-2928.
http://dx.doi.0rg/10.1016/i.atmosenv.2009.03.011
Heck. WW: Cowling. EB. (1997). The need for a long term cumulative secondary ozone standard -
An ecological perspective. EM January: 23-33.
Heck. WW: Cure. WW: Rawlings. JO: Zaragoza. LJ: Heagle. AS: Heggestad. HE: Kohut. RJ: Kress. LW:
Temple. PJ. (1984). Assessing impacts of ozone on agricultural crops: II. Crop yield functions and
alternative exposure statistics. J Air Pollut Control Assoc 34: 810-817.
Heck. WW: Heagle. AS: Miller. JE; Rawlings. JO. (1991). A national program (NCLAN) to assess the impact
of ozone on agricultural resources. In RL Berglund; DR Lawson; DJ McKee (Eds.), Tropospheric ozone
and the environment: papers from an international conference; March 1990; Los Angeles, CA (pp. 225-
254). Pittsburgh, PA: Air & Waste Management Association.
Heck. WW; Philbeck. RB; Dunning. JA. (1978). A continuous stirred tank reactor (CSTR) system for exposing
plants to gaseous air contaminants: Principles, specifications, construction, and operation. Washington, DC:
U.S. Government Printing Office.
Heck. WW; Taylor. PC; Adams. R; Bingham. G; Miller. J; Preston. E; Weinstein. L. (1982). Assessment of
crop loss from ozone. J Air Pollut Control Assoc 32: 353-361.
Heggestad. HE. (1991). Origin of Bel-W3, Bel-C and Bel-B tobacco varieties and their use as indicators of
ozone. Environ Pollut 74: 263-291. http://dx.doi.org/10.1016/0269-7491(91)90076-9
Heidenreich. B; Haberer. G; Mayer. K; Sandermann. H; Ernst. D. (2005). CDNA array-analysis of mercury-
and ozone-induced genes in Arabidopsis thaliana. Acta Physiologiae Plantarum 27: 45-51.
http://dx.doi.org/10.1007/sll738-005-0035-l
Hendrev. GR; Ellsworth. DS; Lewin. KF; Nagv. J. (1999). A free-air enrichment system for exposing tall forest
vegetation to elevated atmospheric CO2. Global Change Biol 5: 293-309. http://dx.doi.Org/10.1046/i. 1365-
2486.1999.00228.x
Hendrev. GR; Kimball. BA. (1994). The FACE program. Agr Forest Meteorol 70: 3-14.
http://dx.doi.org/10.1016/0168-1923(94)90044-2
Hildebrand. E; Skelly. JM; Fredericksen. TS. (1996). Foliar response of ozone-sensitive hardwood tree species
from 1991 to 1993 in the Shenandoah National Park, Virginia. Can J For Res 26: 658-669.
Hillstrom. M; Meehan. TD; Kelly. K; Lindroth. RL. (2010a). Soil carbon and nitrogen mineralization
following deposition of insect frass and greenfall from forests under elevated CO2 and O3. Plant Soil 336:
75-85. http://dx.doi.org/10.1007/slll04-010-0449-4
Hillstrom. ML; Lindroth. RL. (2008). Elevated atmospheric carbon dioxide and ozone alter forest insect
abundance and community composition. Insect Conservation and Diversity 1: 233-241.
http://dx.doi.0rg/10.llll/i.1752-4598.2008.00031.x
Hillstrom. ML; Vigue. LM; Coyle. DR; Raffa. KF; Lindroth. RL. (201 Ob). Performance of the invasive weevil
Polydrusus sericeus is influenced by atmospheric CO2 and host species. Agr Forest Entomol 12: 285-292.
http://dx.doi.0rg/10.llll/i.1461-9563.2010.00474.x
9-163
-------
Himanen. SJ: Nerg. AM; Holopainen. JK. (2009a). Degree of herbivore feeding damage as an important
contributor to multitrophic plant-parasitoid signaling under climate change [Comment]. Plant Signalling &
Behavior 4: 249-251.
Himanen. SJ: Nerg. AM; Nissinen. A; Pinto. DM; Stewart. CN; Poppy. GM; Holopainen. JK. (2009b). Effects
of elevated carbon dioxide and ozone on volatile terpenoid emissions and multitrophic communication of
transgenic insecticidal oilseed rape (Brassica napus). New Phytol 181: 174-186.
http://dx.doi.0rg/10.llll/i.1469-8137.2008.02646.x
Himanen. SJ: Nerg. AM; Nissinen. A; Stewart. CN; Poppy. GM; Holopainen. JK. (2009c). Elevated
atmospheric ozone increases concentration of insecticidal Bacillus thuringiensis (Bt) Cry 1 Ac protein in Bt
Brassica napus and reduces feeding of a Bt target herbivore on the non-transgenic parent. Environ Pollut
157: 181-185. http://dx.doi.0rg/10.1016/i.envpol.2008.07.006
Hogg. A; Uddling. J; Ellsworth. D; Carroll. MA; Presslev. S: Lamb. B; Vogel. C. (2007). Stomatal and non-
stomatal fluxes of ozone to anorthern mixed hardwood forest. Tellus B Chem Phys Meteorol 59: 514-525.
http://dx.doi.0rg/10.llll/i.1600-0889.2007.00269.x
Hogsett. WE; Olszyk. D; Ormrod. DP; Taylor. GE. Jr; Tingev. DT. (1987a). Air pollution exposure systems and
experimental protocols: Volume 2: Description of facilities. (EPA/600/3-87/037b). Corvallis, OR: U.S.
Environmental Protection Agency, Environmental Research Laboratory.
http://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=30000KQH.txt
Hogsett. WE; Olszyk. D; Ormrod. DP; Taylor. GE. Jr; Tingev. DT. (1987b). Air pollution exposure systems
and experimental protocols: Volume I: A review and evaluation of performance. (EPA/600/3-87/037a).
Corvallis, OR: U.S. Environmental Protection Agency.
Hogsett. WE; Tingev. DT: Holman. SR. (1985). A programmable exposure control system for determination of
the effects of pollutant exposure regimes on plant growth. Atmos Environ 19: 1135-1145.
http://dx.doi.org/10.1016/0004-6981(85)90198-2
Hogsett. WE; Tingev. DT: Lee. EH. (1988). Ozone exposure indices: Concepts for development and evaluation
of their use. In Assessment of crop loss from air pollutants: Proceedings of an international conference.
New York: Elsevier Applied Science.
Hogsett. WE; Tingev. DT: Lee. EH; Beedlow. PA; Andersen. CP (2008). An approach for evaluating the
effectiveness of various ozone Air Quality Standards for protecting trees. Environ Manage 41: 937-948.
http://dx.doi.org/10.1007/s00267-007-9057-3
Hogsett. WE; Weber. JE; Tingev. D: Herstrom. A; Lee. EH; Laurence. JA. (1997). Environmental auditing:
An approach for characterizing tropospheric ozone risk to forests. J Environ Manage 21: 105-120.
http://dx.doi.org/10.1007/s002679900010
Holmes. WE; Zak. PR; Pregitzer. KS: King. JS. (2006). Elevated CO2 and O3 alter soil nitrogen
transformations beneath trembling aspen, paper birch, and sugar maple. Ecosystems 9: 1354-1363.
htrp://dx.doi.org/10.1007/sl0021-006-0163-5
Hong. B; Weinstein. D; Swanev. D. (2006). Assessment of ozone effects on nitrate export from Hubbard Brook
Watershed 6. Environ Pollut 141: 8-21. http://dx.doi.Org/10.1016/i.envpol.2005.08.030
Horvath. L; Nagv. Z; Weidinger. T; Artz. R; Luke. WT; Valigura. R; Pinto. JP; Womack. J. (1995).
Measurement of fluxes of trace gases (O3, NOX, SO2, CO2, HNO3), particulate sulfate and nitrate, water
vapour over short vegetation by gradient and eddy correlation techniques in Hungary [Abstract]. Ann
Geophys 13: C490.
Howard. AR; van lersel. MW: Richards. JH; Donovan. LA. (2009). Night-time transpiration can decrease
hydraulic redistribution. Plant Cell Environ 32: 1060-1070. http://dx.doi.org/10.llll/U365-
3040.2009.01988.x
Hui. D; Sims. DA; Johnson. DW: Chang. W: Luo. Y. (2002). Effects of gradual versus step increases in carbon
dioxide on Plantago photosynthesis and growth in a microcosm study. Environ Exp Bot 47: 51-66.
9-164
-------
ICP M&M (International Cooperative Programme on Modelling and Mapping). (2004). Mapping critical levels
for vegetation. In Manual on methodologies and criteria for modelling and mapping critical loads and
levels, and air pollution effects, risks and trends, http://www.rivm.nl/en/themasites/icpmm/manual-and-
downloads/manual-english/index.html
Innes. JL; Skelly. JM; Schaub. M. (2001). Ozon, Laubholz- und Krautpflanzen: Bin Fuhrer zum Bestimmen
von Ozonsymptomen. Bern, Switzerland: Paul Haupt Publishers.
IPCC (Intergovernmental Panel on Climate Change). (2007a). Climate change 2007: Impacts, adaptation and
vulnerability. Cambridge, UK: Cambridge University Press.
Iriti. M; Di Maro. A; Bernasconi. S: Burlini. N; Simonetti. P; Picchi. V; Panigada. C: Gerosa. G: Parente. A;
Faoro. F. (2009). Nutritional traits of bean (Phaseolus vulgaris) seeds from plants chronically exposed to
ozone pollution. J Agric Food Chem 57: 201-208. http://dx.doi.org/10.1021/if802819m
Iriti. M; Faoro. F. (2009). Chemical diversity and defence metabolism: How plants cope with pathogens and
ozone pollution. International Journal of Molecular Sciences 10: 3371-3399.
http://dx.doi.org/10.3390/iimsl0083371
IRRI (International Rice Research Institute). (2002). IRRI annual report 2001-2002. Los Banos, Laguna in the
Philippines, http://www.scribd.com/collections/2641758/Annual-Reports
Isebrands. JG: Dickson. RE; Rebbeck. J; Karnosky. DF (2000). Interacting effects of multiple stresses on
growth and physiological processes in northern forest trees. In RA Mickler; RA Birsdey; J Horn (Eds.),
Responses of northern US forests to environmental change (pp. 149-180). New York, NY: Springer-Verlag.
Isebrands. JG: McDonald. EP; Kruger. E; Hendrev. G: Percy. K; Pregitzer. K; Sober. J: Karnosky. DF. (2001).
Growth responses of Populus tremuloides clones to interacting carbon dioxide and tropospheric ozone.
Environ Pollut 115: 359-371.
Jackson. DM; Heagle. AS: Eckel. RVW. (1999). Ovipositional response of tobacco hornworm moths
(Lepidoptera: Sphyngidae) to tobacco plants grown under elevated levels of ozone. Environ Entomol 28:
566-571.
Janzik. I; Preiskowski. S: Kneifel. H. (2005). Ozone has dramatic effects on the regulation of the
prechorismate pathway in tobacco (Nicotiana tabacum L. cv. Bel W3). Planta 223: 20-27.
http://dx.doi.org/10.1007/s00425-005-0060-8
Johnson. RM; Pregitzer. KS. (2007). Concentration of sugars, phenolic acids, and amino acids in forest soils
exposed to elevated atmospheric CO2 and O3. Soil Biol Biochem 39: 3159-3166.
http://dx.doi.0rg/10.1016/i.soilbio.2007.07.010
Jones. ME; Paine. TD. (2006). Detecting changes in insect herbivore communities along a pollution gradient.
Environ Pollut 143: 377-387. http://dx.doi.org/10.1016/i.envpol.20Q5.12.013
Jones. MLM; Hodges. G: Mills. G. (2010). Nitrogen mediates above-ground effects of ozone but not below-
ground effects in a rhizomatous sedge. Environ Pollut 158: 559-565.
http://dx.doi.0rg/10.1016/i.envpol.2009.08.002
Jones. TG: Freeman. C: Lloyd. A; Mills. G. (2009). Impacts of elevated atmospheric ozone on peatland below-
ground doc characteristics. Ecol Eng 35: 971-977. http://dx.doi.Org/10.1016/j.ecoleng.2008.08.009
Joo. JH; Wang. SY; Chen. JG: Jones. AM; Fedoroff. NV. (2005). Different signaling and cell death roles of
heterotrimeric G protein alpha and beta subunits in the arabidopsis oxidative stress response to ozone. Plant
Cell 17: 957-970. http://dx.doi.org/10.1105/tpc.104.029603
Joss. U: Graber. WK. (1996). Profiles and simulated exchange of H2O, O3, NO2 between the atmosphere and
the HartX Scots pine plantation. Theor Appl Climatol 53: 157-172.
Jovan. S: McCune. B. (2006). Using epiphytic macrolichen communities for biomonitoring ammonia in forests
of the greater Sierra Nevada, California. Water Air Soil Pollut 170: 69-93.
9-165
-------
Kanerva. T; Palojarvi. A; Ramo. K; Manninen. S. (2008). Changes in soil microbial community structure under
elevated tropospheric O3 and CO2. Soil Biol Biochem 40: 2502-2510.
http://dx.doi.0rg/10.1016/i.soilbio.2008.06.007
Kanerva. T; Palojarvi. A; Ramo. K; Ojanpera. K; Esala. M; Manninen. S. (2006). A 3-year exposure to CO2
and O3 induced minor changes in soil N cycling in a meadow ecosystem. Plant Soil 286: 61-73.
http://dx.doi.org/10.1007/slll04-006-9026-2
Kanerva. T; Regina. K; Ramo. K; Ojanpera. K; Manninen. S. (2007). Fluxes of N2O, CH4 and CO2 in a
meadow ecosystem exposed to elevated ozone and carbon dioxide for three years. Environ Pollut 145: 818-
828. http://dx.doi.0rg/10.1016/j.envpol.2006.03.055
Kangasjarvi. J; Jaspers. P; Kollist. H. (2005). Signalling and cell death in ozone-exposed plants. Plant Cell
Environ 28: 1021-1036.
Karlsson. PE; Sellden. G: Plaijel. H. (2003). Establishing ozone critical levels II: UNECE workshop report.
Gothenburg, Sweden: IVL Swedish Environmental Institute.
Karlsson. PE; Uddling. J; Braun. S: Broadmeadow. M; Elvira. S: Gimeno. BS: Le Thiec. D; Okansen. E;
Vandermeiren. K; Wilkinson. M; Emberson. L. (2004). New critical levels for ozone effects on young trees
based on AOT40 and simulated cumulative leaf uptake of ozone. Atmos Environ 38: 2283-2294.
Karnosky. DF; Gagnon. ZE; Dickson. RE; Coleman. MD; Lee. EH; Isebrands. JG. (1996). Changes in growth,
leaf abscission, biomass associated with seasonal tropospheric ozone exposures of Populus tremuloides
clones and seedlings. Can J For Res 26: 23-37.
Karnosky. DF; Mankovska. B; Percy. K; Dickson. RE; Podila. GK; Sober. J: Noormets. A; Hendrev. G:
Coleman. MD; Kubiske. M; Pregitzer. KS: Isebrands. JG. (1999). Effects of tropospheric ozone on
trembling aspen and interaction with CO2: Results from an O3-gradient and a FACE experiment. Water Air
Soil Pollut 116: 311-322.
Karnosky. DF; Pregitzer. KS: Zak. PR; Kubiske. ME; Hendrev. GR; Weinstein. D; Nosal. M; Percy. KE.
(2005). Scaling ozone responses of forest trees to the ecosystem level in a changing climate. Plant Cell
Environ 28: 965-981. http://dx.doi.Org/10.llll/i.1365-3040.2005.01362.x
Karnosky. DF; Zak. PR; Pregitzer. KS: Awmack. CS: Bockheim. JG: Dickson. RE; Hendrev. GR: Host. GE;
King. JS: Kopper. BJ; Kruger. EL; Kubiske. ME; Lindroth. RL; Mattson. WJ: McDonald. EP; Noormets.
A; Oksanen. E; Parsons. WFJ; Percy. KE; Podila. GK; Riemenschneider. DE; Sharma. P; Thakur. R; Sober.
A; Sober. J: Jones. WS: Anttonen. S: Vapaavuori. E; Mankovska. B; Heilman. W: Isebrands. JG. (2003).
Tropospheric O3 moderates responses of temperate hardwood forests to elevated CO2: A synthesis of
molecular to ecosystem results from the Aspen FACE project. Funct Ecol 17: 289-304.
Kasurinen. A; Keinanen. MM; Kaipainen. S: Nilsson. LO; Vapaavuori. E; Kontro. MH; Holopainen. T. (2005).
Below-ground responses of silver birch trees exposed to elevated CO2 and O3 levels during three growing
seasons. Global Change Biol 11: 1167-1179. http://dx.doi.Org/10.llll/i.1365-2486.2005.00970.x
Kasurinen. A; Peltonen. PA; Julkunen-Tiitto. R; Vapaavuori. E; Nuutinen. V; Holopainen. T; Holopainen. JK.
(2007). Effects of elevated CO2 and O3 on leaf litter phenolics and subsequent performance of litter-
feeding soil macrofauna. Plant Soil 292: 25-43. http://dx.doi.org/10.1007/slll04-007-9199-3
Kasurinen. A; Riikonen. J: Oksanen. E; Vapaavuori. E; Holopainen. T. (2006). Chemical composition and
decomposition of silver birch leaf litter produced under elevated CO2 and O3. Plant Soil 282: 261-280.
http://dx.doi.org/10.1007/slll04-005-6026-6
Kats. G: Olszyk. DM; Thompson. CR. (1985). Open top experimental chambers for trees. J Air Waste Manag
Assoc35: 1298-1301.
Kats. G: Thompson. CR: Kubv. WC. (1976). Improved ventilation of open top greenhouses. J Air Pollut
Control Assoc 26: 1089-1090.
Keller. T; Ha'sler. R. (1984). The influence of a fall fumigation with ozone on the stomatal behavior of spruce
and fir. Oecologia 64: 284-286. http://dx.doi.org/10.1007/BF00376884
9-166
-------
Kellomaki. S: Wang. KY. (1997). Effects of elevated O3 and CO2 concentrations on photosynthesis and
stomatal conductance in Scots pine. Plant Cell Environ 20: 995-1006. http://dx.doi.org/10.1111/i. 1365-
3040.1997.tb00676.x
Kerner. R; Winkler. J; Dupuy. J; Jiirgensen. M; Lindermayr. C: Ernst. D; Miiller-starck. G. (2011). Changes in
the proteome of juvenile European beech following three years exposure to free-air elevated ozone. iForest
4: 69-76. http://dx.doi.org/10.3832/ifor0570-004
Keutgen. AJ; Noga. G: Pawelzik. E. (2005). Cultivar-specific impairment of strawberry growth,
photosynthesis, carbohydrate and nitrogen accumulation by ozone. Environ Exp Bot 53: 271-280.
http://dx.doi.0rg/10.1016/i.envexpbot.2004.04.003
Keutgen. N; Keutgen. AJ: Janssens. MJJ. (2008). Sweet potato [Ipomoea batatas (L.) Lam.] cultivated as tuber
or leafy vegetable supplier as affected by elevated tropospheric ozone. J Agric Food Chem 56: 6686-6690.
http://dx.doi.org/10.1021/if8006272
King. JS: Kubiske. ME; Pregitzer. KS: Hendrev. GR: McDonald. EP; Giardina. CP; Quinn. VS: Karnosky. DF.
(2005). Tropospheric O3 compromises net primary production in young stands of trembling aspen, paper
birch and sugar maple in response to elevated atmospheric CO2. New Phytol 168: 623-635.
http://dx.doi.org/10.1111/i.1469-8137.2005.01557.x
King. JS: Pregitzer. KS: Zak. PR; Sober. J: Isebrands. JG: Dickson. RE; Hendrev. GR: Karnosky. DF. (2001).
Fine-root biomass and fluxes of soil carbon in young stands of paper birch and trembling aspen as affected
by elevated atmospheric CO2 and tropospheric O3. Oecologia 128: 237-250.
Kitao. M; Low. M; Heerdt C: Grams. TEE; Haberle. KH; Matvssek. R. (2009). Effects of chronic elevated
ozone exposure on gas exchange responses of adult beech trees (Fagus sylvatica) as related to the within-
canopy light gradient. Environ Pollut 157: 537-544. http://dx.doi.Org/10.1016/i.envpol.2008.09.016
Kline. LJ; Davis. DP; Skelly. JM; Decoteau. DR. (2009). Variation in ozone sensitivity within Indian hemp
and common milkweed selections from the Midwest. Northeast Nat 16: 307-313.
http://dx.doi.org/10.1656/045.016.0210
Kline. LJ: Davis. DP; Skelly. JM; Savage. JE; Ferdinand. J. (2008). Ozone sensitivity of 28 plant selections
exposed to ozone under controlled conditions. Northeast Nat 15: 57-66. http://dx.doi.org/10.1656/1092-
6194(2008)15r57:OSOPSE12.0.CO:2
Kohut. R. (2007). Assessing the risk of foliar injury from ozone on vegetation in parks in the US National Park
Service's Vital Signs Network. Environ Pollut 149: 348-357.
http://dx.doi.0rg/10.1016/i.envpol.2007.04.022
Kolb. TE; Fredericksen. TS: Steiner. KG: Skelly. JM. (1997). Issues in scaling tree size and age responses to
ozone: A review [Review]. Environ Pollut 98: 195-208. http://dx.doi.org/10.1016/S0269-7491(97)00132-2
Kollist. T; Moldau. H; Rasulov. B; Oja. V: Ramma. H; Huve. K; Jaspers. P: Kangasiarvi. J: Kollist. H. (2007).
A novel device detects a rapid ozone-induced transient stomatal closure in intact Arabidopsis and its
absence in abi2 mutant. Physiol Plant 129: 796-803. http://dx.doi.Org/10.llll/j.1399-3054.2006.00851.x
Kostka-Rick. R; Halm. HU. (2005). Biomonitoring using tobacco Bel W3 provides supplemental information
for risk assessment of vegetation injury due to ozone. Gefahrstoffe Reinhaltung Der Luft 65: 485-491.
Kozovits. AR; Matvssek. R; Blaschke. H; Gottlein. A; Grams. TEE. (2005). Competition increasingly
dominates the responsiveness of juvenile beech and spruce to elevated CO2 and/or O3 concentrations
throughout two subsequent growing seasons. Global Change Biol 11: 1387-1401.
http://dx.doi.0rg/10.llll/i.1365-2486.2005.00993.x
Krupa. SV: Grunhage. L; Jager. HJ; Nosal. M; Manning. WJ: Legge. AH; Hanewald. K. (1995). Ambient
ozone (O3) and adverse crop response: Aunified view of cause and effect. Environ Pollut 87: 119-126.
http://dx.doi.org/10.1016/S0269-7491(99)80014-l
Krupa. SV: Nosal. M; Peterson. PL. (2001). Use of passive ozone O3 samplers in vegetation effects
assessment. Environ Pollut 112: 303-309.
9-167
-------
Kubiske. ME; Quinn. VS: Heilman. WE; McDonald. EP; Marquardt. PE; Teclaw. RM; Friend. AL; Karnoskev.
DF. (2006). Interannual climatic variation mediates elevated CO2 and O3 effects on forest growth. Global
Change Biol 12: 1054-1068. http://dx.doi.Org/10.llll/i.1365-2486.2006.01152.x
Kubiske. ME; Quinn. VS: Marquardt. PE; Karnosky. DF. (2007). Effects of elevated atmospheric CO2 and/or
O3 on intra- and interspecific competitive ability of aspen. Plant Biol (Stuttg) 9: 342-355.
http://dx.doi.org/10.1055/s-2006-924760
Kurpius. MR; Goldstein. AH. (2003). Gas-phase chemistry dominates O3 loss to a forest, implying a source of
aerosols and hydroxyl radicals to the atmosphere. Geophys Res Lett 30.
Laffrav. X; Rose. C: Garrec. JP. (2007). Estimation of ozone concentration in a valley of the alps mountains
based on bel-w3 tobacco leaf injury. Water Air Soil Pollut 186: 29-42. http://dx.doi.org/10.1007/sll270-
007-9460-7
Lambers. H; Chapin. FS. Ill; Pons. TL. (1998). Plant physiological ecology. In H Lambers; FS Chapin, III; TL
Pons (Eds.). New York: Springer.
Langebartels. C: Kerner. K; Leonardi. S: Schraudner. M; Trost. M; Heller. W: Sandermann. H. Jr. (1991).
Biochemical plant responses to ozone: I. Differential induction of polyamine and ethylene biosynthesis in
tobacco. J Plant Physiol 95: 882-889. http://dx.doi.Org/10.1104/pp.95.3.882
Larson. JL; Zak. PR; Sinsabaugh. RL. (2002). Extracellular enzyme activity beneath temperate trees growing
under elevated carbon dioxide and ozone. Soil Sci Soc Am J 66: 1848-1856.
Lawlor. DW. (1998). Plant responses to global change: Temperature and drought stress. In LJ De Kok; I Stolen
(Eds.), Responses of plant metabolism to air pollution and global change. Leiden, The Netherlands:
Backhuys Publishers.
Leakey. ADB; Bernacchi. CJ: Ort. PR; Long. SP (2006). Long-term growth of soybean at elevated CO2 does
not cause acclimation of stomatal conductance under fully open-air conditions. Plant Cell Environ 29:
1794-1800. http://dx.doi.0rg/10.llll/j.1365-3040.2006.01556.x
Lee. EH; Hogsett. WE. (1996). Methodology for calculating inputs for ozone secondary standard benefits
anaylsis: Part II. Research Triangle Park, NC: U.S. Environmental Protection Agency.
Lee. EH; Hogsett. WE. (1999). Role of concentrations and time of day in developing ozone exposure indices
for a secondary standard. J Air Waste Manag Assoc 49: 669-681.
Lee. EH; Hogsett. WE; Tingev. DT. (1994). Attainment and effects issues regarding alternative secondary
ozone air quality standards. J Environ Qual 23: 1129-1140.
http://dx.doi.org/10.2134/ieal994.00472425002300060002x
Lee. EH; Tingev. DT; Hogsett. WE. (1987). Selection of the best exposure-response model using various 7-
hour ozone exposure statistics. Research Triangle Park, NC: U.S. Environmental Protection Agency.
Lee. EH; Tingev. DT; Hogsett. WE. (1988a). Evaluation of ozone-exposure indices for relating exposure to
plant production and for estimating agricultural losses. (EPA/600/3-88/039). Washington, DC: U.S.
Environmental Protection Agency.
Lee. EH; Tingev. DT; Hogsett. WE. (1988b). Evaluation of ozone exposure indices in exposure-response
modeling. Environ Pollut 53: 43-62. http://dx.doi.org/10.1016/0269-7491(88)90024-3
Lee. EH; Tingev. DT; Hogsett. WE. (1989). Interrelation of experimental exposure and ambient air quality data
for comparison of ozone exposure indices and estimating agricultural losses. (EPA/600/3-89/047).
Corvallis, OR: U.S. Environmental Protection Agency.
Lee. EH; Tingev. DT; Hogsett. WE; Laurence. JA. (2003a). History of tropospheric ozone for the San
Bernardino Mountains of southern California, 1963-1999. Atmos Environ 37: 2705-2717.
http://dx.doi.org/10.1016/81352-2310(03)00203-6
Lee. EH; Tingev. DT: Waschmann. RS: Phillips. PL; Olszyk. DM; Johnson. MG: Hogsett. WE. (2009a).
Seasonal and long-term effects of CO2 and O3 on water loss in ponderosa pine and their interaction with
climate and soil moisture. Tree Physiol 29: 1381-1393. http://dx.doi.org/10.1093/treephys/tpp071
9-168
-------
Lee. S: Yun. SC. (2006). The ozone stress transcriptome of pepper (Capsicum annuum L.). Molecules and
Cells 21: 197-205.
Lee. WS: Chevone. BI; Seller. JR. (1990). Growth and gas exchange of loblolly pine seedlings as influenced
by drought and air pollutants. Water Air Soil Pollut 51: 105-116. http://dx.doi.org/10.1007/BF00211508
Lefohn. AS; Benedict. HM. (1982). Development of a mathematical index that describes ozone concentration,
frequency and duration. Atmos Environ 16: 2529-2532. http://dx.doi.org/10.1016/0004-6981(82)90145-7
Lefohn. AS: Jackson. W: Shadwick. DS; Knudsen. HP. (1997). Effect of surface ozone exposures on
vegetation grown in the southern Appalachian Mountains: Identification of possible areas of concern.
Atmos Environ 31: 1695-1708. http://dx.doi.org/10.1016/81352-2310(96)00258-0
Lefohn. AS: Laurence. JA; Kohut. RJ. (1988). A comparison of indices that describe the relationship between
exposure to ozone and reduction in the yield of agricultural crops. Atmos Environ 22: 1229-1240.
http://dx.doi.org/10.1016/0004-6981(88)90353-8
Lefohn. AS: Runeckles. VC. (1987). Establishing standards to protect vegetation - ozone exposure/dose
considerations. Atmos Environ 21: 561-568. http://dx.doi.org/10.1016/0004-6981(87)90038-2
Lefohn. AS: Shadwick. DS. (2000). Differences in trending estimates in the United States using several ozone
metrics. In Proceedings of the 93rd Air & Waste Management Association Annual Conference and
Exhibition. Pittsburgh, PA: Air & Waste Management Association.
Legge. AH; Grunhage. L; Nosal. M; Jager. HJ; Krupa. SV. (1995). Ambient ozone and adverse crop response:
An evaluation of North American and European data as they relate to exposure indices and critical levels. J
Appl Hot Food Qua! 69: 192-205.
Leitao. L; Bethenod. O: Biollev. JP (2007a). The impact of ozone on juvenile maize (Zeamays L.) plant
photosynthesis: Effects on vegetative biomass, pigmentation, and carboxylases (PEPc and Rubisco). Plant
Biol (Stuttg) 9: 478-488. http://dx.doi.org/10.1055/s-2007-964942
Leitao. L; Delacote. E; Dizengremel. P; Le Thiec. D; Biollev. JP. (2007b). Assessment of the impact of
increasing concentrations of ozone on photosynthetic components of maize (Zeamays L.), a C-4 plant.
Environ Pollut 146: 5-8. http://dx.doi.Org/10.1016/i.envpol.2006.05.019
Leitao. L; Maoret. JJ: Biollev. JP. (2007c). Changes in PEP carboxylase, rubisco and rubisco activase mRNA
levels from maize (Zea mays) exposed to a chronic ozone stress. Biol Res 40: 137-153.
http://dx.doi.org/10.4067/S0716-97602007000200005
Lesser. VM; Rawlings. JO: Spruill. SE; Somerville. MC. (1990). Ozone effects on agricultural crops:
Statistical methodologies and estimated dose-response relationships. Crop Sci 30: 148-155.
Leuning. R; Unsworth. MH; Neumann. HN; King. KM. (1979). Ozone fluxes to tobacco and soil under field
conditions. Atmos Environ 13: 1155-1163. http://dx.doi.org/10.1016/0004-6981(79)90039-8
Levine. JS: Pinto. JP. (1998). The production of CO by biomass burning. In MAK Khalil; JP Pinto; MJ Shearer
(Eds.), Atmospheric carbon monoxide and its environmental effects: Proceedings of the international
conference; December 1997; Portland, Oregon (pp. 251-256). Research Triangle Park, NC: U.S.
Environmental Protection Agency, Office of Research and Development.
Lewis. JS; Ditchkoff. SS; Lin. JC; Muntifering. RB; Chappelka. AH. (2006). Nutritive quality of big bluestem
(Andropogon gerardii) and eastern gamagrass (Tripsacum dactyloides) exposed to tropospheric ozone.
Rangeland Ecol Manag 59: 267-274.
Li. PH; Mane. SP; Sioson. AA; Robinet. CV; Heath. LS; Bohnert. HJ; Grene. R. (2006b). Effects of chronic
ozone exposure on gene expression in Arabidopsis thaliana ecotypes and in Thellungielia halophila. Plant
Cell Environ 29: 854-868. http://dx.doi.Org/10.llll/i.1365-3040.2005.01465.x
Lin. JC; Nosal. M; Muntifering. RB; Krupa. SV. (2007). Alfalfa nutritive quality for ruminant livestock as
influenced by ambient air quality in west-central Alberta. Environ Pollut 149: 99-103.
http://dx.doi.0rg/10.1016/i.envpol.2006.12.009
9-169
-------
Lindroth. RL. (2010). Impacts of elevated atmospheric CO2 and O3 on forests: Phytochemistry, trophic
interactions, and ecosystem dynamics [Review]. J Chem Ecol 36: 21-Feb.
http://dx.doi.org/10.1007/sl0886-009-9731-4
Liu. L; King. J; Giardina. C. (2005). Effects of elevated concentrations of atmospheric CO2 and tropospheric
O3 on leaf litter production and chemistry in trembling aspen and paper birch communities. Tree Physiol
25: 1511-1522.
Liu. L; King. JS: Giardina. CP. (2007a). Effects of elevated atmospheric CO2 and tropospheric O3 on nutrient
dynamics: Decomposition of leaf litter in trembling aspen and paper birch communities. Plant Soil 299: 65-
82. http://dx.doi.org/10.1007/slll04-007-9361-v
Liu. LL; King. JS: Giardina. CP: Booker. FL. (2009b). The influence of chemistry, production and community
composition on leaf litter decomposition under elevated atmospheric CO2 and tropospheric O-3 in a
northern hardwood ecosystem. Ecosystems 12: 401-416. http://dx.doi.org/10.1007/sl0021-009-9231-y
Loats. KV; Rebbeck. J. (1999). Interactive effects of ozone and elevated carbon dioxide on the growth and
physiology of black cherry, green ash, and yellow poplar seedlings. Environ Pollut 106: 237-248.
http://dx.doi.org/10.1016/S0269-7491(99)00069-X
Long. SP (1991). Modification of the response of photo synthetic productivity to rising temperature by
atmospheric CO2 concentrations: Has its importance been underestimated? Plant Cell Environ 14: 729-739.
http://dx.doi.0rg/10.llll/i.1365-3040.1991.tb01439.x
Loranger. GI; Pregitzer. KS: King. JS. (2004). Elevated CO2 and O3t concentrations differentially affect
selected groups of the fauna in temperate forest soils. Soil Biol Biochem 36: 1521-1524.
Lorenzini. G: Nali. C. (1995). Analysis of vertical ozone and nitrogen oxides profiles in a Primus cerasifera
canopy. Int J Biometeorol 39: 1-4. http://dx.doi.org/10.1007/BF01320885
Loreto. F; Fares. S. (2007). Is ozone flux inside leaves only a damage indicator? Clues from volatile isoprenoid
studies. Plant Physiol 143: 1096-1100. http://dx.doi.org/10.1104/pp.106.091892
Loreto. F; Velikova. V. (2001). Isoprene produced by leaves protects the photosynthetic apparatus against
ozone damage, quenches ozone products, and reduces lipid peroxidation of cellular membranes. Plant
Physiol 127: 1781-1787. http://dx.doi.org/10.1104/pp.010497
Low. M; Herbinger. K; Nunn. AJ; Haberle. KH; Leuchner. M; Heerdt. C: Werner. H; Wipfler. P; Pretzsch. H;
Tausz. M; Matvssek. R. (2006). Extraordinary drought of 2003 overrules ozone impact on adult beech trees
(Fagus sylvatica). Trees Struct Funct 20: 539-548. http://dx.doi.org/10.1007/s00468-006-0069-z
Loya. WM; Pregitzer. KS: Karberg. NJ; King. JS: Giardina. CP. (2003). Reduction of soil carbon formation by
tropospheric ozone under elevated carbon dioxide. Nature 425: 705-707.
Ludwikow. A; Gallois. P; Sadowski. J. (2004). Ozone-induced oxidative stress response in Arabidopsis:
Transcription profiling by microarray approach. Cell Mol Biol Lett 9: 829-842.
Ludwikow. A; Kierzek. D; Gallois. P: Zeef. L; Sadowski. J. (2009). Gene expression profiling of ozone-treated
Arabidopsis abiltd insertional mutant: Protein phosphatase 2C ABU modulates biosynthesis ratio of ABA
and ethylene. Planta 230: 1003-1017. http://dx.doi.org/10.1007/s00425-009-1001-8
Ludwikow. A; Sadowski. J. (2008). Gene networks in plant ozone stress response and tolerance. J Integr Plant
Biol 50: 1256-1267. http://dx.doi.Org/10.llll/i.1744-7909.2008.00738.x
Luo. Y. (2001). Transient ecosystem response to free-air CO2 enrichment (FACE): Experimental evidence and
methods of analysis. New Phytol 152: 3-8.
Luo. Y; Reynolds. JF (1999). Validity of extrapolating field CO2 experiments to predict carbon sequestration
in natural ecosystems. Ecology 80: 1568-1583.
Lyons. TM; Barnes. JD. (1998). Influence of plant age on ozone resistance in Plantago major. New Phytol 138:
83-89. http://dx.doi.0rg/10.1046/i.1469-8137.1998.00879.x
9-170
-------
Maggio. A; Chiaranda. FQ; Cefariello. R; Fagnano. M. (2009). Responses to ozone pollution of alfalfa exposed
to increasing salinity levels. Environ Pollut 157: 1445-1452. http://dx.doi.Org/10.1016/i.envpol.2008.09.013
Mahalingam. R; Jambunathan. N; Gunjan. SK; Faustin. E; Weng. H; Ayoubi. P. (2006). Analysis of oxidative
signalling induced by ozone in Arabidopsis thaliana. Plant Cell Environ 29: 1357-1371.
http://dx.doi.0rg/10.llll/i.1365-3040.2006.01516.x
Mahalingam. R; Shah. N; Scrvmgeour. A; Fedoroff. N. (2005). Temporal evolution of the Arabidopsis
oxidative stress response. Plant Mol Biol 57: 709-730. http://dx.doi.org/10.1007/slll03-005-2860-4
Maier-Maercker. U. (1998). Predisposition of trees to drought stress by ozone. Tree Physiol 19: 71-78.
Mandl. RH; Laurence. JA; Kohut. RJ. (1989). Development and testing of open-top chambers for exposing
large, perennial plants to air pollutants. J Environ Qual 18: 534-540.
http://dx.doi.org/10.2134/ieal989.00472425001800040026x
Mandl. RH; Weinstein. LH; McCune. DC: Keveny. M. (1973). A cylindrical, open-top chamber for the
exposure of plants to air pollutants in the field. J Environ Qual 2: 371-376.
Mankovska. B; Percy. KE; Karnosky. DF. (2005). Impacts of greenhouse gases on epicuticular waxes of
Populus tremuloides Michx.: Results from an open-air exposure and a natural O3 gradient. Environ Pollut
137: 580-586. http://dx.doi.Org/10.1016/i.envpol.2005.01.043
Manning. WJ. (2003). Detecting plant effects is necessary to give biological significance to ambient ozone
monitoring data and predictive ozone standards. Environ Pollut 126: 375-379.
Manning. WJ: Krupa. SV. (1992). Experimental methodology for studying the effects of ozone on crops and
trees. In AS Lefohn (Ed.), Surface level ozone exposures and their effects on vegetation (pp. 93-156).
Chelsea, MI: Lewis Publishers.
Martin. MJ; Host. GE; Lenz. KE; Isebrands. JG. (2001). Simulating the growth response of aspen to elevated
ozone: A mechanistic approach to scaling a leaf-level model of ozone effects on photosynthesis to a
complex canopy architecture. Environ Pollut 115: 425-436.
Massman. WJ. (2004). Toward an ozone standard to protect vegetation based on effective dose: A review of
deposition resistances and a possible metric [Review]. Atmos Environ 38: 2323-2337.
Massman. WJ: Grantz. DA. (1995). Estimating canopy conductance to ozone uptake from observations of
evapotranspiration at the canopy scale and at the leaf scale. Global Change Biol 1: 183-198.
http://dx.doi.org/10.1111/i. 1365-2486.1995.tb00020.x
Massman. WJ: Musselman. RC: Lefohn. AS. (2000). A conceptual ozone dose-response model to develop a
standard to protect vegetation. Atmos Environ 34: 745-759. http://dx.doi.org/10.1016/S1352-
2310(99)00395-7
Matvssek. R; Gunthardt-Goerg. MS: Maurer. S: Keller. T. (1995). Nighttime exposure to ozone reduces whole-
plant production in Betulapendula. Tree Physiol 15: 159-165.
Matvssek. R; Le Thiec. D; Low. M; Dizengremel. P; Nunn. AJ; Haberle. KH. (2006). Interactions between
drought and O3 stress in forest trees. Plant Biol (Stuttg) 8: 11-17. http://dx.doi.org/10.1055/s-2005-873025
Matvssek. R; Sandermann. H; Wieser. G: Booker. F; Cieslik. S: Musselman. R; Ernst. D. (2008).
The challenge of making ozone risk assessment for forest trees more mechanistic. Environ Pollut 156: 567-
582. http://dx.doi.0rg/10.1016/j.envpol.2008.04.017
Matvssek. R; Wieser. G: Ceulemans. R: Rennenberg. H; Pretzsch. H; Haberer. K; Low. M; Nunn. AJ: Werner.
H; Wipfler. P; Obwald. W: Nikolova. P; Hanke. DE; Kraigher. H; Tausz. M; Bahnweg. G: Kitao. M; Dieler.
J; Sandermann. H; Herbinger. K; Grebenc. T; Blumenrother. M; Deckmyn. G: Grams. TEE; Heerdt. C:
Leuchner. M; Fabian. P; Haberle. KH. (2010). Enhanced ozone strongly reduces carbon sink strength of
adult beech (Fagus sylvatica): Resume from the free-air fumigation study at Kranzberg Forest. Environ
Pollut 158: 2527-2532. http://dx.doi.Org/10.1016/i.envpol.2010.05.009
Mautz. WJ: Dohm. MR. (2004). Respiratory and behavioral effects of ozone on a lizard and a frog. Comp
Biochem Physiol A Mol Integr Physiol 139: 371-377. http://dx.doi.Org/10.1016/i.cbpb.2004.10.004
9-171
-------
McAinsh. MR; Evans. NH; Montgomery. LT; North. KA. (2002). Calcium signalling in stomatal responses to
pollutants. New Phytol 153: 441-447.
McBride. JR: Laven. RD. (1999). Impact of oxidant air pollutants on forest succession in the mixed conifer
forests of the San Bernardino Mountains. In PR Miller; JR McBride (Eds.), Oxidant air pollution impacts in
the montane forests of southern California: A case study of the San Bernardino Mountains (pp. 338-352).
New York, NY: Springer-Verlag.
McCarthy. HR; Oren. R; Johnsen. KH; Gallet-Budynek. A; Pritchard. SG: Cook. CW: LaDeau. SL; Jackson.
RB; Finzi. AC. (2009). Re-assessment of plant carbon dynamics at the Duke free-air CO2 enrichment site:
Interactions of atmospheric [CO2] with nitrogen and water availability over stand development. New
Phytol 185: 514-528. htrp://dx.doi.org/10.1111/i.l469-8137.2009.03078.x
McCool. PM; Musselman. RC: Younglove. T; Teso. RR. (1988). Response of kidney bean to sequential ozone
exposures. Environ Exp Bot 28: 307-313.
McFrederick. QS: Fuentes. JD; Roulston. T; Kathilankal. JC: Lerdau. M. (2009). Effects of air pollution on
biogenic volatiles and ecological interactions. Oecologia 160: 411-420. http://dx.doi.org/10.1007/s00442-
009-1318-9
McFrederick. QS: Kathilankal. JC: Fuentes. JD. (2008). Air pollution modifies floral scent trails. Atmos
Environ 42: 2336-2348. http://dx.doi.Org/10.1016/i.atmosenv.2007.12.033
McLaughlin. SB; Nosal. M; Wullschleger. SD; Sun. G. (2007a). Interactive effects of ozone and climate on
tree growth and water use in a southern Appalachian forest in the USA. New Phytol 174: 109-124.
http://dx.doi.0rg/10.llll/i.1469-8137.2007.02018.x
McLaughlin. SB; Wullschleger. SD: Sun. G: Nosal. M. (2007b). Interactive effects of ozone and climate on
water use, soil moisture content and streamflow in a southern Appalachian forest in the USA. New Phytol
174: 125-136. http://dx.doi.0rg/10.llll/i.1469-8137.2007.01970.x
McLeod. AR; Long. SP (1999). Free-air carbon dioxide enrichment (FACE) in global change research:
A review. Adv Ecol Res 28: 1-56. http://dx.doi.org/10.1016/S0065-2504(08)60028-8
Medlyn. BE; Barton. CVM; Broadmeadow. MSJ: Ceulemans. R; De Angelis. P; Forstreuter. M; Freeman. M;
Jackson. SB; Kellomaki. S: Laitat. E; Rev. A; Roberntz. P; Sigurdsson. BD; Strassemever. J; Wang. K;
Curtis. PS: Jarvis. PG. (2001). Stomatal conductance of forest species after long-term exposure to elevated
CO2 concentration: A synthesis. New Phytol 149: 247-264. http://dx.doi.org/10.1046/U469-
8137.2001.00028.x
Meehan. TD; Crosslev. MS: Lindroth. RL. (2010). Impacts of elevated CO2 and O3 on aspen leaf litter
chemistry and earthworm and springtail productivity. Soil Biol Biochem 42: 1132-1137.
htto://dx.doi.org/10.1016/i.soilbio.2010.03.019
Menendez. AI; Romero. AM; Folcia. AM; Martinez-Ghersa. MA. (2009). Getting the interactions right: Will
higher O3 levels interfere with induced defenses to aphid feeding? Basic Appl Ecol 10: 255-264.
http://dx.doi.0rg/10.1016/i.baae.2008.03.010
Menendez. AI; Romero. AM; Folcia. AM; Martinez-Ghersa. MA. (2010). Aphid and episodic O3 injury in
arugula plants (Eruca sativa Mill) grown in open-top field chambers. Agric Ecosyst Environ 135: 10-14.
http://dx.doi.0rg/10.1016/i.agee.2009.08.005
Mereu. S: Gerosa. G: Finco. A; Fusaro. L; Muys. B; Manes. F. (2009). Improved sapflow methodology reveals
considerable night-time ozone uptake by Mediterranean species. Biogeosciences 6: 3151-3162.
http://dx.doi.0rg/10.5194/bg-6-3151-2009
Miles. GP; Samuel. MA; Zhang. YL; Ellis. BE. (2005). RNA interference-based (RNAi) suppression of
AtMPK6, an Arabidopsis mitogen-activated protein kinase, results in hypersensitivity to ozone and
misregulation of AtMPK3. Environ Pollut 138: 230-237. http://dx.doi.Org/10.1016/i.envpol.2005.04.017
Miller. PL. (1973). Oxidant-induced community change in a mixed conifer forest. In JANaegele (Ed.), Air
pollution damage to vegetation (pp. 101-117). Washington, DC: American Chemical Society.
9-172
-------
Miller. PR; McCutchan. MH; Ryan. BC. (1972). Influence of climate and topography on oxidant air pollution
concentrations that damage conifer forests in southern California. Mitt Forstl Bundesversuchsanst Wien 97:
585-607.
Miller. PR; Parmeter. JR. Jr; Taylor. PC: Cardiff. EA. (1963). Ozone injury to the foliage of Pinus ponderosa.
Phytopathology 53: 1072-1076.
Miller. PR; Rechel. J. (1999). Temporal changes in crown condition indices, needle litterfall, and collateral
needle injuries of Ponderosa and Jeffrey pines. In PR Miller; JR McBride (Eds.), Oxidant air pollution
impacts in the montane forests of southern California: A case study of the San Bernardino Mountains (pp.
164-178). New York, NY: Springer.
Mills. G. (2002). Modification of plant response by environmental conditions. In JNB Bell; M Treshow (Eds.),
Air pollution and plant life (2nd ed., pp. 343-358). Chichester, United Kingdom: John Wiley & Sons.
Mills. G; Ball. G; Haves. F; Fuhrer. J; Skarbv. L; Gimeno. B; De Temmerman. L. (2000). Development of a
multi-factor model for predicting the effects of ambient ozone on the biomass of white clover. Environ
Pollut 109: 533-542. http://dx.doi.org/10.1016/S0269-7491(00)00057-9
Mills. G; Buse. A; Gimeno. B; Bermejo. V; Holland. M; Emberson. L; Pleijel. H. (2007a). A synthesis of
AOT40-based response functions and critical levels of ozone for agricultural and horticultural crops. Atmos
Environ 41: 2630-2643. http://dx.doi.Org/10.1016/i.atmosenv.2006.ll.016
Mills. G; Haves. F; Jones. MLM; Cinderbv. S. (2007b). Identifying ozone-sensitive communities of (semi-
)natural vegetation suitable for mapping exceedance of critical levels. Environ Pollut 146: 736-743.
http://dx.doi.0rg/10.1016/i.envpol.2006.04.005
Mills. G; Haves. F; Wilkinson. S; Davies. WJ. (2009). Chronic exposure to increasing background ozone
impairs stomatal functioning in grassland species. Global Change Biol 15: 1522-1533.
http://dx.doi.0rg/10.llll/i.1365-2486.2008.01798.x
Mondor. EB; Awmack. CS; Lindroth. RL. (2010). Individual growth rates do not predict aphid population
densities under altered atmospheric conditions. Agr Forest Entomol 12: 293-299.
http://dx.doi.0rg/10.llll/i.1461-9563.2010.00478.x
Mondor. EB; Tremblav. MN; Awmack. CS; Lindroth. RL. (2004). Divergent pheromone-mediated insect
behaviour under global atmospheric change. Global Change Biol 10: 1820-1824.
Mondor. EB; Tremblav. MN; Awmack. CS; Lindroth. RL. (2005). Altered genotypic and phenotypic
frequencies of aphid populations under enriched CO2 and O3 atmospheres. Global Change Biol 11: 1990-
1996. http://dx.doi.0rg/10.llll/i.1365-2486.2005.01054.x
Morgan. PB; Ainsworth. EA; Long. SP (2003). How does elevated ozone impact soybean? A meta-analysis of
photosynthesis, growth and yield. Plant Cell Environ 26: 1317-1328.
Morgan. PB; Bernacchi. CJ; Ort. PR; Long. SP. (2004). An in vivo analysis of the effect of season-long open-
air elevation of ozone to anticipated 2050 levels on photosynthesis in soybean. J Plant Physiol 135: 2348-
2357.
Morgan. PB; Mies. TA; Bollero. GA; Nelson. RL; Long. SP. (2006). Season-long elevation of ozone
concentration to projected 2050 levels under fully open-air conditions substantially decreases the growth
and production of soybean. New Phytol 170: 333-343. http://dx.doi.Org/10.llll/i.1469-8137.2006.01679.x
Morison. JIL; Lawlor. DW. (1999). Interactions between increasing CO2 concentration and temperature on
plant growth. Plant Cell Environ 22: 659-682. http://dx.doi.Org/10.1046/i.1365-3040.1999.00443.x
Morsky. SK; Haapala. JK; Rinnan. R; Tiiva. P; Saarnio. S; Silvola. J; Holopainen. T; Martikainen. PJ. (2008).
Long-term ozone effects on vegetation, microbial community and methane dynamics of boreal peatland
microcosms in open-field conditions. Global Change Biol 14: 1891-1903. http://dx.doi.org/10.Ill 1/j. 1365-
2486.2008.01615.x
Mudd. JB. (1996). Biochemical basis for the toxicity of ozone. In M Yunus; M Iqbal (Eds.), Plant response to
air pollution (pp. 267-283). New York, NY: John Wiley & Sons.
9-173
-------
Muntifering. RB: Chappelka. AH; Lin. JC: Karnosky. DF; Somers. GL. (2006a). Chemical composition and
digestibility of Trifolium exposed to elevated ozone and carbon dioxide in a free-air (FACE) fumigation
system. Funct Ecol 20: 269-275. http://dx.doi.Org/10.llll/i.1365-2435.2006.01093.x
Muntifering. RB: Manning. WJ: Lin. JC: Robinson. GB. (2006b). Short-term exposure to ozone altered the
relative feed value of an alfalfa cultivar. Environ Pollut 140: 1-3.
Musselman. RC: Lefohn. AS: Massman. WJ: Heath. RL. (2006). A critical review and analysis of the use of
exposure- and flux-based ozone indices for predicting vegetation effects [Review]. Atmos Environ 40:
1869-1888. http://dx.doi.0rg/10.1016/i.atmosenv.2005.10.064
Musselman. RC: Massman. WJ. (1999). Ozone flux to vegetation and its relationship to plant response and
ambient air quality standards. Atmos Environ 33: 65-73.
Musselman. RC: McCool. PM; Younglove. T. (1988). Selecting ozone exposure statistics for determining crop
yield loss from air pollutants. Environ Pollut 53: 63-78. http://dx.doi.org/10.1016/0269-7491(88)90025-5
Musselman. RC: Minnick. TJ. (2000). Nocturnal stomatal conductance and ambient air quality standards for
ozone. Atmos Environ 34: 719-733. http://dx.doi.org/10.1016/S1352-2310(99)00355-6
Nali. C: Balducci. E; Frati. L; Paoli. L; Loppi. S: Lorenzini. G. (2007). Integrated biomonitoring of air quality
with plants and lichens: A case study on ambient ozone from central Italy. Chemosphere 67: 2169-2176.
http://dx.doi.0rg/10.1016/i.chemosphere.2006.12.036
NAPAP (National Acid Precipitation Assessment Program). (1987). Diagnosing injury to Eastern forest trees:
A manual for identifying damage caused by air pollution, pathogens, insects, and abiotic stresses.
University Park, PA: Pennsylvania State University.
Nash. TH. III. (2008). Lichen sensitivity to air pollution. In TH Nash, III (Ed.), Lichen biology (2nd ed., pp.
301-316). Cambridge, UK: Cambridge University Press.
Neufeld. HS: Lee. EH; Renfro. JR; Hacker. WD. (2000). Seedling insensitivity to ozone for three conifer
species native to Great Smoky Mountains National Park. Environ Pollut 108: 141-151.
http://dx.doi.org/10.1016/S0269-7491(99)00247-X
Neufeld. HS: Lee. EH; Renfro. JR; Hacker. WD: Yu. BH. (1995). Sensitivity of seedlings of black cherry
(Prunus serotina Ehrh) to ozone in Great Smoky Mountains National Park I Exposure-response curves for
biomass. New Phytol 130: 447-459. http://dx.doi.Org/10.llll/i.1469-8137.1995.tb01839.x
Neufeld. HS: Renfro. JR; Hacker. WD: Silsbee. D. (1992). Ozone in Great Smoky Mountains National Park:
Dynamics and effects on plants. In RL Berglund (Ed.), Tropospheric ozone and the environment II: Effects,
modeling and control (pp. 594-617). Pittsburgh, PA: Air & Waste Management Association.
Nikolova. PS: Andersen. CP; Blaschke. H; Matvssek. R; Haberle. KH. (2010). Belowground effects of
enhanced tropospheric ozone and drought in a beech/spruce forest (Fagus sylvatica L./Picea abies [L.]
Karst). Environ Pollut 158: 1071-1078. http://dx.doi.Org/10.1016/j.envpol.2009.07.036
Noctor. G: Foyer. CH. (1998). Ascorbate and glutathione: Keeping active oxygen under control. Annu Rev
Plant Physiol Plant Mol Biol 49: 249-279. http://dx.doi.Org/10.1146/annurev.arplant.49.l.249
Norbv. RJ: DeLucia. EH; Gielen. B; Calfapietra. C: Giardina. CP: King. JS: Ledford. J: McCarthy. HR; Moore.
DJP; Ceulemans. R; De Angelis. P; Finzi. AC: Karnosky. DF; Kubiske. ME; Lukac. M; Pregitzer. KS:
Scarascia-Mugnozza. GE; Schlesinger. WH; Oren. R. (2005). Forest response to elevated CO2 is conserved
across a broad range of productivity. PNAS 102: 18052-18056. http://dx.doi.org/10.1073/pnas.0509478102
Novak. K; Cherubini. P; Saurer. M; Fuhrer. J; Skelly. JM; Krauchi. N; Schaub. M. (2007). Ozone air pollution
effects on tree-ring growth, delta(13)C, visible foliar injury and leaf gas exchange in three ozone-sensitive
woody plant species. Tree Physiol 27: 941-949.
NFS (U.S. National Park Service). (2006). Ozone bioindicators. Washington, DC.
http://www.nature.nps.gov/air/Pubs/bioindicators/index.cfm
NFS (U.S. National Park Service). (2007). Ozone effects studies. Washington, DC: U.S. Department of the
Interior, National Park Service. http://www.nature.nps.gov/air/studies/ecoOzone.cfm
9-174
-------
Nunn. AJ; Reiter. IM; Haberle. KH; Werner. H; Langebartels. C: Sandermann. H; Heerdt. C: Fabian. P;
Matvssek. R. (2002). Free-air ozone canopy fumigation in an old-growth mixed forest: concept and
observations in beech. Phyton (Buenos Aires) 42: 105-119.
Nussbaum. S: Geissmann. M; Fuhrer. J. (1995). Ozone exposure-response relationships for mixtures of
perennial ryegrass and white clover depend on ozone exposure patterns. Atmos Environ 29: 989-995.
http://dx.doi.org/10.1016/1352-2310(94)00368-U
O'Gara. PJ. (1922). Sulfur dioxide and fume problems and their solutions [Abstract]. The Journal of Industrial
and Engineering Chemistry 14: 744.
O'Neill. BF; Zangerl. AR; Delucia. EH; Berenbaum. MR. (2008). Longevity and fecundity of Japanese beetle
(Popillia japonica) on foliage grown under elevated carbon dioxide. Environ Entomol 37: 601-607.
O'Neill. BF; Zangerl. AR; Dermodv. O: Bilgin. DP; Casteel. CL; Zavala. JA; DeLucia. EH; Berenbaum. MR.
(2010). Impact of elevated levels of atmospheric CO2 and herbivory on flavonoids of soybean (Glycine
max Linnaeus). J Chem Ecol 36: 35-45. http://dx.doi.org/10.1007/sl0886-009-9727-0
Ogawa. D; Nakajima. N; Sano. T; Tamaoki. M; Aono. M; Kubo. A; Kanna. M; loki. M; Kamada. H; Saji. H.
(2005). Salicylic acid accumulation under O-3 exposure is regulated by ethylene in tobacco plants. Plant
Cell Physiol 46: 1062-1072. http://dx.doi.org/10.1093/pcp/pcill8
Oksanen. E. (2003). Responses of selected birch (Betula pendula Roth) clones to ozone change over time.
Plant Cell Environ 26: 875-886.
Oksanen. E; Holopainen. T. (2001). Responses of two birch (Betula pendula Roth) clones to different ozone
profiles with similar AOT40 exposure. Atmos Environ 35: 5245-5254. http://dx.doi.org/10.1016/S1352-
2310(01)00346-6
Olbrich. M; Betz. G: Gerstner. E; Langebartels. C: Sandermann. H; Ernst. D. (2005). Transcriptome analysis of
ozone-responsive genes in leaves of European beech (Fagus sylvatica L.). Plant Biol (Stuttg) 7: 670-676.
http://dx.doi.org/10.1055/s-2005-873001
Olbrich. M: Gerstner. E: Bahnweg. G: Haberle. KH: Matvssek. R: Welzl. G: Heller. W: Ernst. D. (2010).
Transcriptional signatures in leaves of adult European beech trees (Fagus sylvatica L.) in an experimentally
enhanced free air ozone setting. Environ Pollut 158: 977-982.
http://dx.doi.0rg/10.1016/i.envpol.2009.08.001
Olbrich. M; Gerstner. E; Welzl. G: Winkler. JB; Ernst. D. (2009). Transcript responses in leaves of ozone-
treated beech saplings seasons at an outdoor free air model fumigation site over two growing seasons. Plant
Soil 323: 61-74. http://dx.doi.org/10.1007/slll04-009-0129-4
Ollinger. SV: Aber. JD; Reich. PB. (1997a). Simulating ozone effects on forest productivity: Interactions
among leaf- and stand-level processes. Ecol Appl 123: 351-358.
Ollinger. SV: Aber. JD; Reich. PB. (1997b). Simulating ozone effects on forest productivity: Interactions
among leaf-, canopy-, and stand-level processes. Ecol Appl 7: 1237-1251.
Ollinger. SV: Aber. JD: Reich. PB: Freuder. RJ. (2002). Interactive effects of nitrogen deposition, tropospheric
ozone, elevated CO2 and land use history on the carbon dynamics of northern hardwood forests. Global
Change Biol 8: 545-562. http://dx.doi.Org/10.1046/j.1365-2486.2002.00482.x
Olszyk. DM; Kats. G: Dawson. PJ: Bytnerowicz. A; Wolf. J: Thompson. CR. (1986). Characteristics of air
exclusion systems vs chambers for field air pollution studies. J Environ Qua! 15: 326-334.
Omasa. K; Shimazaki. KI; Aiga. I; Larcher. W: Onoe. M. (1987). Image analysis of chlorophyll fluorescence
transients for diagnosing the photosynthetic system of attached leaves. Plant Physiol 84: 748-752.
httrj://dx.doi.org/10.1104/pp.84.3.748
Orendovici-Best T; Skelly. JM; Davis. DP; Ferdinand. JA; Savage. JE; Stevenson. RE. (2008). Ozone uptake
(flux) as it relates to ozone-induced foliar symptoms of Prunus serotina and Populus maximowizii x
trichocarpa. Environ Pollut 151: 79-92. http://dx.doi.Org/10.1016/i.envpol.2007.03.003
9-175
-------
Orendovici. T; Skelly. JM; Ferdinand. JA; Savage. JE; Sanz. MJ; Smith. GC. (2003). Response of native plants
of northeastern United States and southern Spain to ozone exposures; determining exposure/response
relationships. Environ Pollut 125: 31-40.
Oshima. RJ: Braegelmann. PK: Baldwin. DW: Van Way. V: Taylor. PC. (1977). Reduction of tomato fruit size
and yield by ozone. J Am Soc Hortic Sci 102: 289-293.
Oshima. RJ: Poe. MP; Braegelmann. PK; Baldwin. DW: Van Way. V. (1976). Ozone dosage-crop loss function
for alfalfa: A standardized method for assessing crop losses from air pollutants. J Air Pollut Control Assoc
26: 861-865.
Overmyer. K; Brosche. M; Pellinen. R; Kuittinen. T; Tuominen. H; Ahlfors. R; Keinanen. M; Saarma. M;
Scheel. D; Kangasjarvi. J. (2005). Ozone-induced programmed cell death in the Arabidopsis radical-
induced cell deathl mutant. Plant Physiol 137: 1092-1104. http://dx.doi.org/10.1104/pp.104.055681
Overmyer. K; Kollist. H; Tuominen. H; Betz. C: Langebartels. C: Wingsle. G: Kangasjarvi. S: Brader. G:
Mullineaux. P; Kangasjarvi. J. (2008). Complex phenotypic profiles leading to ozone sensitivity in
Arabidopsis thaliana mutants. Plant Cell Environ 31: 1237-1249. http://dx.doi.org/10.llll/U365-
3040.2008.01837.x
Overmyer. K; Tuominen. H; Kettunen. R; Betz. C: Langebartels. C: Sandermann. H. Jr; Kangasjarvi. J. (2000).
Ozone-sensitive Arabidopsis rcdl mutant reveals opposite roles for ethylene and jasmonate signaling
pathways in regulating superoxide-dependent cell death. Plant Cell 12: 1849-1862.
http://dx.doi.org/10.1105/tpc.12.10.1849
Pan. YD; Birdsev. R: Horn. J: McCullough. K. (2009). Separating effects of changes in atmospheric
composition, climate and land-use on carbon sequestration of US Mid-Atlantic temperate forests. For Ecol
Manage 259: 151-164. http://dx.doi.Org/10.1016/i.foreco.2009.09.049
Panek. J: Kurpius. MR; Goldstein. AH. (2002). An evaluation of ozone exposure metrics for a seasonally
drought-stressed ponderosa pine ecosystem. Environ Pollut 117: 93-100. http://dx.doi.org/10.1016/S0269-
7491(01)00155-5
Panek. JA. (2004). Ozone uptake, water loss and carbon exchange dynamics in annually drought-stressed
Pinus ponderosa forests: Measured trends and parameters for uptake modeling. Tree Physiol 24: 277-290.
Panek. JA: Goldstein. AH. (2001). Responses of stomatal conductance to drought in ponderosa pine:
Implications for carbon and ozone uptake. Tree Physiol 21: 337-344.
Paolacci. AR; Miraldi. C: Tanzarella. OA; Badiani. M; Porceddu. E; Nali. C: Lorenzini. G: Ciaffi. M. (2007).
Gene expression induced by chronic ozone in the Mediterranean shrub Phillyrea latifolia: Analysis by
cDNA-AFLP. Tree Physiol 27: 1541-1550. http://dx.doi.org/10.1093/treephys/27.11.1541
Paoletti. E; Grulke. NE. (2005). Does living in elevated CO2 ameliorate tree response to ozone? A review on
stomatal responses [Review]. Environ Pollut 137: 483-493. http://dx.doi.Org/10.1016/i.envpol.2005.01.035
Paoletti. E; Grulke. NE. (2010). Ozone exposure and stomatal sluggishness in different plant physiognomic
classes. Environ Pollut 158: 2664-2671. http://dx.doi.Org/10.1016/i.envpol.2010.04.024
Paoletti. E; Manning. WJ. (2007). Toward a biologically significant and usable standard for ozone that will
also protect plants. Environ Pollut 150: 85-95. http://dx.doi.Org/10.1016/i.envpol.2007.06.037
Paoletti. E; Seufert. G: Delia Rocca. G: Thomsen. H. (2007). Photosynthetic responses to elevated CO2 and O-
3 in Quercus ilex leaves at a natural CO2 spring. Environ Pollut 147: 516-524.
http://dx.doi.0rg/10.1016/i.envpol.2006.08.039
Parsons. WFJ; Bockheim. JG: Lindroth. RL. (2008). Independent, interactive, and species-specific responses
of leaf litter decomposition to elevated CO2 and O3 in a northern hardwood forest. Ecosystems 11: 505-
519.
Pearson. M; Mansfield. TA. (1993). Interacting effects of ozone and water stress on the stomatal resistance of
beech (Fagus sylvatica L). New Phytol 123: 351-358. http://dx.doi.org/10.1111/i. 1469-
8137.1993.tb03745.x
9-176
-------
Pearson. S: Davison. AW: Reiling. K; Ashenden. T; Ollerenshaw. JH. (1996). The effects of different ozone
exposures on three contrasting populations of Plantago major. New Phytol 132: 493-502.
http://dx.doi.0rg/10.llll/i.1469-8137.1996.tb01869.x
Pei. ZM; Murata. Y; Benning. G: Thomine. S: Kliisener. B; Allen. GJ: Grill. E; Schroeder. JI. (2000). Calcium
channels activated by hydrogen peroxide mediate abscisic acid signalling in guard cells. Nature 406: 731-
734. http://dx.doi.org/10.1038/35021067
Pell. EJ; Sinn. JP; Brendlev. BW: Samuelson. L; Vinten-Johansen. C: Tien. M; Skillman. J. (1999). Differential
response of four tree species to ozone-induced acceleration of foliar senescence. Plant Cell Environ 22:
779-790. http://dx.doi.0rg/10.1046/i.1365-3040.1999.00449.x
Peltonen. PA; Julkunen-Tiitto. R; Vapaavuori. E; Holopainen. JK. (2006). Effects of elevated carbon dioxide
and ozone on aphid oviposition preference and birch bud exudate phenolics. Global Change Biol 12: 1670-
1679. http://dx.doi.0rg/10.llll/i.1365-2486.2006.01226.x
Peltonen. PA; Vapaavuori. E; Heinonen. J; Julkunen-Tiitto. R; Holopainen. JK. (2010). Do elevated
atmospheric CO2 and O-3 affect food quality and performance of folivorous insects on silver birch? Global
Change Biol 16: 918-935. http://dx.doi.Org/10.llll/i.1365-2486.2009.02073.x
Percy. KE; Nosal. M; Heilman. W: Dann. T; Sober. J: Legge. AH; Karnosky. DF. (2007). New exposure-based
metric approach for evaluating O3 risk to North American aspen forests. Environ Pollut 147: 554-566.
http://dx.doi.0rg/10.1016/i.envpol.2006.10.009
Peterson. PL; Arbaugh. MJ; Wakefield. VA; Miller. PR. (1987). Evidence of growth reduction in ozone-injured
Jeffrey pine (Pinus jeffreyi Grev and Balf) in Sequoia and Kings Canyon National Parks. J Air Waste
Manag Assoc 37: 906-912.
Pfleeger. TG: Plocher. M; Bichel. P. (2010). Response of pioneer plant communities to elevated ozone
exposure. Agric Ecosyst Environ 138: 116-126. http://dx.doi.Org/10.1016/i.agee.2010.04.009
Phillips. PL; Johnson. MG: Tingev. DT; Storm. MJ. (2009). Elevated CO2 and O3 effects on fine-root
survivorship in ponderosa pine mesocosms. Oecologia 160: 827-837. http://dx.doi.org/10.1007/s00442-
009-1339-4
Phillips. RL; Zak. PR; Holmes. WE; White. DC. (2002). Microbial community composition and function
beneath temperate trees exposed to elevated atmospheric carbon dioxide and ozone. Oecologia 131: 236-
244.
Piikki. K; De Temmerman. L; Hogy. P; Pleijel. H. (2008a). The open-top chamber impact on vapour pressure
deficit and its consequences for stomatal ozone uptake. Atmos Environ 42: 6513-6522.
http://dx.doi.0rg/10.1016/i.atmosenv.2008.04.014
Piikki. K; De Temmerman. L; Ojanpera. K; Danielsson. H; Pleijel. H. (2008b). The grain quality of spring
wheat (Triticum aestivum L.) in relation to elevated ozone uptake and carbon dioxide exposure. Eur J
Agron 28: 245-254. http://dx.doi.Org/10.1016/i.eia.2007.07.004
Piikki. K; Vorne. V: Ojanpera. K; Pleijel. H. (2007). Impact of elevated O3 and CO2 exposure on potato
(Solanum tuberosum L. cv. Bintje) tuber macronutrients (N, P, K, Mg, Ca). Agric Ecosyst Environ 118: 55-
64. http://dx.doi.0rg/10.1016/i.agee.2006.04.012
Pinto. DM; Blande. JD; Nykanen. R; Dong. WX; Nerg. AM; Holopainen. JK. (2007a). Ozone degrades
common herbivore-induced plant volatiles: Does this affect herbivore prey location by predators and
parasitoids? J Chem Ecol 33: 683-694. http://dx.doi.org/10.1007/sl0886-007-9255-8
Pinto. DM; Blande. JD: Souza. SR; Nerg. AM; Holopainen. JK. (2010). Plant volatile organic compounds
(VOCs) in ozone (O3) polluted atmospheres: The ecological effects [Review]. J Chem Ecol 36: 22-34.
http://dx.doi.org/10.1007/sl0886-009-9732-3
Pinto. DM; Himanen. SJ: Nissinen. A; Nerg. AM; Holopainen. JK. (2008). Host location behavior of Cotesia
plutellae Kurdjumov (Hymenoptera: Braconidae) in ambient and moderately elevated ozone in field
conditions. Environ Pollut 156: 227-231. http://dx.doi.Org/10.1016/i.envpol.2007.12.009
9-177
-------
Pinto. DM; Nerg. AM; Holopainen. JK. (2007b). The role of ozone-reactive compounds, terpenes, and green
leaf volatiles (GLVs), in the orientation of Cotesia plutellae. J Chem Ecol 33: 2218-2228.
http://dx.doi.org/10.1007/sl0886-007-9376-0
Pinto. J. (2009). Wyoming winter smog. Nat Geosci 2: 88-90. http://dx.doi.org/10.1038/ngeo430
Pleijel. H; Danielsson. H; Gelang. J; Sild. E; Sellden. G. (1998). Growth stage dependence of the grain yield
response to ozone in spring wheat (Triticum aestivum L). Agric Ecosyst Environ 70: 61-68.
http://dx.doi.org/10.1016/S0167-8809(97)00167-9
Pleijel. H; Danielsson. H; Ojanpera. K; De Temmerman. L; Hogy. P; Badiani. M; Karlsson. PE. (2004a).
Relationships between ozone exposure and yield loss in European wheat and potato~a comparison of
concentration- and flux-based exposure indices. Atmos Environ 38: 2259-2269.
Pleijel. H; h. D; Ojanpera. K; De Temmerman. L; Hogy. P. (2004b). Relationships between ozone exposure and
yield loss in wheat and potato - Suggestions of critical levels for ozone effects on crops. Atmos Environ 38:
2259-2269. http://dx.doi.Org/10.1016/i.atmosenv.2003.09.076
Pleijel. H; Ojanpera. K; Mortensen. L. (1997). Effects of tropospheric ozone on the yield and grain protein
content of spring wheat (Triticum aestivum L) in the nordic countries. Acta Agric Scand B Soil Plant Sci
47: 20-25. http://dx.doi.org/10.1080/09064719709362434
Plessl. M; Elstner. EF; Rennenberg. H; Habermever. J: Heiser. I. (2007). Influence of elevated CO2 and ozone
concentrations on late blight resistance and growth of potato plants. Environ Exp Bot 60: 447-457.
http://dx.doi.0rg/10.1016/i.envexpbot.2007.01.003
Plochl. M; Lyons. T; Ollerenshaw. J: Barnes. J. (2000). Simulating ozone detoxification in the leaf apoplast
through the direct reaction with ascorbate. Planta 210: 454-467. http://dx.doi.org/10.1007/PL00008153
Pollastrini. M; Desotgiu. R; Cascio. C: Bussotti. F; Cherubini. P; Saurer. M; Gerosa. G: Marzuoli. R. (2010).
Growth and physiological responses to ozone and mild drought stress of tree species with different
ecological requirements. Trees Struct Funct 24: 695-704. http://dx.doi.org/10.1007/s00468-010-0439-4
Polle. A; Pell. EJ. (1999). Role of carbon dioxide in modifying the plant response to ozone. In Y Luo; HA
Mooney (Eds.), Carbon dioxide and environmental stress (pp. 193-213). San Diego, CA: Academic Press.
Pregitzer. K; Loya. W: Kubiske. M; Zak. D. (2006). Soil respiration in northern forests exposed to elevated
atmospheric carbon dioxide and ozone. Oecologia 148: 503-516. http://dx.doi.org/10.1007/s00442-006-
0381-8
Pregitzer. KS: Burton. AJ; King. JS: Zak. DR. (2008). Soil respiration, root biomass, and root turnover
following long-term exposure of northern forests to elevated atmospheric Co-2 and tropospheric O-3. New
Phytol 180: 153-161. http://dx.doi.Org/10.llll/i.1469-8137.2008.02564.x
Pretzsch. H; Dieler. J: Matvssek. R; Wipfler. P. (2010). Tree and stand growth of mature Norway spruce and
European beech under long-term ozone fumigation. Environ Pollut 158: 1061-1070.
http://dx.doi.0rg/10.1016/i.envpol.2009.07.035
Pritsch. K; Esperschuetz. J: Haesler. F; Raidl. S: Winkler. B; Schloter. M. (2009). Structure and activities of
ectomycorrhizal and microbial communities in the rhizosphere of Fagus sylvatica under ozone and
pathogen stress in a lysimeter study. Plant Soil 323: 97-109. http://dx.doi.org/10.1007/slll04-009-9972-6
Puckette. MC: Tang. YH; Mahalingam. R. (2008). Transcriptomic changes induced by acute ozone in resistant
and sensitive Medicago truncatula accessions. BMC Plant Biol 8: 46. http://dx.doi.org/10.1186/1471-2229-
8-46
Pujol Pereira. El; Chung. H; Scow. K; Sadowsky. MJ; van Kessel. C: Six. J. (2011). Soil nitrogen
transformations under elevated atmospheric CO and O during the soybean growing season. Environ Pollut
159: 401-407. http://dx.doi.0rg/10.1016/i.envpol.2010.10.033
Ramo. K; Kanerva. T; Ojanpera. K; Manninen. S. (2007). Growth onset, senescence, and reproductive
development of meadow species in mesocosms exposed to elevated O3 and CO2. Environ Pollut 145: 850-
860. http://dx.doi.0rg/10.1016/i.envpol.2006.03.054
9-178
-------
Rao. MV; Hale. BA; Ormrod. DP. (1995). Amelioration of ozone-induced oxidative damage in wheat plants
grown under high carbon dioxide: Role of antioxidant enzymes. J Plant Physiol 109: 421-432.
Rawlings. JO: Cure. WW. (1985). The Weibull function as a dose-response model to describe ozone effects on
crop yields. Crop Sci 25: 807-814.
Reich. PB. (1987). Quantifying plant response to ozone: A unifying theory. Tree Physiol 3: 63-91.
http://dx.doi.0rg/10.1093/treephys/3.l.63
Reich. PB; Lassoie. JP (1984). Effects of low level O3 exposure on leaf diffusive conductance and water-use
efficiency in hybrid poplar. Plant Cell Environ 7: 661-668. http://dx.doi.org/10.1111/1365-
304Q.epll571645
Reid. CD; Fiscus. EL. (2008). Ozone and density affect the response of biomass and seed yield to elevated
CO2 in rice. Global Change Biol 14: 60-76. http://dx.doi.Org/10.llll/i.1365-2486.2007.01472.x
Reiling. K; Davison. AW. (1992). Effects of a short ozone exposure given at different stages in the
development of Plantago major L. New Phytol 121: 643-647. http://dx.doi.org/10.Ill 1/i. 1469-
8137.1992.tb01135.x
Reiling. K; Davison. AW. (1994). Effects of exposure to ozone at different stages in the development of
Plantago major L on chlorophyll fluorescence and gas exchange. New Phytol 128: 509-514.
http://dx.doi.0rg/10.llll/i.1469-8137.1994.tb02998.x
Reinert. RA; Eason. G: Barton. J. (1997). Growth and fruiting of tomato as influenced by elevated carbon
dioxide and ozone. New Phytol 137: 411-420. http://dx.doi.Org/10.1046/i.1469-8137.1997.00846.x
Reinert. RA; Ho. MC. (1995). Vegetative growth of soybean as affected by elevated carbon dioxide and ozone.
Environ Pollut 89: 89-96. http://dx.doi.org/10.1016/0269-7491(94)00039-G
Ren. W: Tian. H; Chen. G: Liu. M; Zhang. C: Chappelka. AH; Pan. S. (2007a). Influence of ozone pollution
and climate variability on net primary productivity and carbon storage in China's grassland ecosystems
from 1961 to 2000. Environ Pollut 149: 327-335. http://dx.doi.Org/10.1016/i.envpol.2007.05.029
Ren. W: Tian. H; Tao. B; Chappelka. A; Sun. G: Lu. C: Liu. M; Chen. G: Xu. X. (2011). Impacts of
tropospheric ozone and climate change on net primary productivity and net carbon exchange of China's
forest ecosystems. Global Ecology and Biogeography 20: 391-406. http://dx.doi.org/10.1111/i. 1466-
8238.2010.00606.x
Ren. W: Tian. HQ; Liu. ML; Zhang. C: Chen. GS: Pan. SF; Felzer. B; Xu. XF. (2007b). Effects of tropospheric
ozone pollution on net primary productivity and carbon storage in terrestrial ecosystems of China. J
Geophys Res 112: D22S09. http://dx.doi.org/10.1029/2007id008521
Rhea. L; King. J: Kubiske. M; Saliendra. N; Teclaw. R. (2010). Effects of elevated atmospheric CO2 and
tropospheric O3 on tree branch growth and implications for hydrologic budgeting. Environ Pollut 158:
1079-1087. http://dx.doi.0rg/10.1016/i.envpol.2009.08.038
Riddell. J: Padgett. PE; Nash. TH. III. (2010). Responses of the lichen Ramalina menziesii Tayl. to ozone
fumigations. In TH Nash, III (Ed.), Biology of lichens: Symbiosis, ecology, environmental monitoring,
systematics, cyber applications (pp. 113-123). Stuttgart: J. Cramer in der Gebriider Borntraeger
Verlagsbuchhandlung.
Riikonen. J: Kets. K; Darbah. J: Oksanen. E; Sober. A; Vapaavuori. E; Kubiske. ME; Nelson. N; Karnosky. DF.
(2008). Carbon gain and bud physiology in Populus tremuloides and Betula papyrifera grown under long-
term exposure to elevated concentrations of CO2 and O3. Tree Physiol 28: 243-254.
http://dx.doi.Org/10.1093/treephys/28.2.243
Riikonen. J: Maenpaa. M; Alavillamo. M; Silfver. T; Oksanen. E. (2009). Interactive effect of elevated
temperature and O3 on antioxidant capacity and gas exchange in Betula pendula saplings. Planta 230: 419-
427. http://dx.doi.org/10.1007/s00425-009-0957-8
Rizzo. M; Bernardi. R; Salvini. M; Nali. C: Lorenzini. G: Durante. M. (2007). Identification of differentially
expressed genes induced by ozone stress in sensitive and tolerant poplar hybrids. J Plant Physiol 164: 945-
949. http://dx.doi.0rg/10.1016/i.iplph.2006.07.012
9-179
-------
Rodenkirchen. H; Gottlein. A; Kozovits. AR; Matvssek. R; Grams. TEE. (2009). Nutrient contents and
efficiencies of beech and spruce saplings as influenced by competition and O3/CO2 regime. European
Journal of Forest Research 128: 117-128. http://dx.doi.org/10.1007/sl0342-008-0221-y
Rogers. A; Allen. DJ; Davev. PA; Morgan. PB; Ainsworth. EA; Bernacchi. CJ: Cornic. G: Dermodv. OC:
Dohleman. FG: Heaton. EA; Mahonev. J: Zhu. XG: Delucia. EH; Ort. PR; Long. SP. (2004). Leaf
photosynthesis and carbohydrate dynamics of soybean grown throughout their life-cycle under free-air
carbon dioxide enrichment. Plant Cell Environ 27: 449-458.
Rowland-Bamford. AJ. (2000). Plant responses to changing carbon dioxide and temperature. In SN Singh
(Ed.), Trace gas emissions and plants (pp. 63-74). Dordecht, The Netherlands: Kluwer Academic
Publishers.
Ryan. A; Cojocariu. C: Possell. M; Davies. WJ: Hewitt. CN. (2009). Defining hybrid poplar (Populus deltoides
x Populus trichocarpa) tolerance to ozone: Identifying key parameters. Plant Cell Environ 32: 31-45.
http://dx.doi.0rg/10.llll/i.1365-3040.2008.01897.x
Rvang. SZ: Woo. SY: Kwon. SY: Kim. SH: Lee. SH: Kim. KN: Lee. DK. (2009). Changes of net
photosynthesis, antioxidant enzyme activities, and antioxidant contents of Liriodendron tulipifera under
elevated ozone. Photosynthetica 47: 19-25. http://dx.doi.org/10.1007/sll099-009-0005-8
Samuel. MA; Ellis. BE. (2002). Double jeopardy: Both overexpression and suppression of a redox-activated
plant mitogen-activated protein kinase render tobacco plants ozone sensitive. Plant Cell 14: 2059-2069.
http://dx.doi.org/10.1105/tpc.002337
Samuel. MA; Miles. GP; Ellis. BE. (2000). Ozone treatment rapidly activates MAP kinase signalling in plants.
Plant J 22: 367-376. http://dx.doi.Org/10.1046/i.1365-313x.2000.00741.x
Samuel. MA; Walia. A; Mansfield. SD; Ellis. BE. (2005). Overexpression of SIPK in tobacco enhances ozone-
induced ethylene formation and blocks ozone-induced SA accumulation. J Exp Bot 56: 2195-2201.
http://dx.doi.org/10.1093/ixb/eri219
Samuelson. LJ; Kelly. JM. (1997). Ozone uptake in Prunus serotina, Acer rubrum and Quercus rubra forest
trees of different sizes. New Phytol 136: 255-264. http://dx.doi.Org/10.1046/j.1469-8137.1997.00734.x
Sanchez. MJS: Pena. GS: Lorente. VC: Gallego. TM; Albert. JC. (2001). La contamination atmosferica en los
bosques: Guia para la identificacion de danos visibles causados por Ozono. Madrid, Spain: Ministerio de
Medio Ambiente. http://www.ceam.es/VentaLibros/guia O3/index.htm
Sandermann. H. (2008). Ecotoxicology of ozone: Bioactivation of extracellular ascorbate. Biochem Biophys
Res Commun 366: 271-274. http://dx.doi.Org/10.1016/j.bbrc.2007.12.018
Sanz. J; Bermejo. V: Gimeno. BS: Elvira. S: Alonso. R. (2007). Ozone sensitivity of the Mediterranean
terophyte trifolium striatum is modulated by soil nitrogen content. Atmos Environ 41: 8952-8962.
http://dx.doi.0rg/10.1016/i.atmosenv.2007.08.016
Sanz. J; Muntifering. RB; Bermejo. V: Gimeno. BS: Elvira. S. (2005). Ozone and increased nitrogen supply
effects on the yield and nutritive quality of Trifolium subterraneum. Atmos Environ 39: 5899-5907.
http://dx.doi.0rg/10.1016/i.atmosenv.2005.06.022
Sarkar. A; Rakwal. R: Agrawal. SB; Shibato. J: Ogawa. Y; Yoshida. Y; Agrawal. GK; Agrawal. M. (2010).
Investigating the impact of elevated levels of ozone on tropical wheat using integrated phenotypical,
physiological, biochemical, and proteomics approaches. J Proteome Res 9: 4565-4584.
http://dx.doi.org/10.1021/Prl002824
Saviranta. NMM; Julkunen-Tiitto. R: Oksanen. E; Karjalainen. RO. (2010). Leaf phenolic compounds in red
clover (Trifolium pratense L.) induced by exposure to moderately elevated ozone. Environ Pollut 158: 440-
446. http://dx.doi.0rg/10.1016/j.envpol.2009.08.029
Sawada. H; Kohno. Y. (2009). Differential ozone sensitivity of rice cultivars as indicated by visible injury and
grain yield. Plant Biol (Stuttg) 11: 70-75. http://dx.doi.Org/10.llll/i.1438-8677.2009.00233.x
9-180
-------
Scebba. F; Giuntini. D; Castagna. A; Soldatini. G: Ranieri. A. (2006). Analysing the impact of ozone on
biochemical and physiological variables in plant species belonging to natural ecosystems. Environ Exp Bot
57: 89-97. http://dx.doi.Org/10.1016/i.envexpbot.2005.04.005
Schaub. M; Skelly. JM; Zhang. JW: Ferdinand. JA; Savage. JE; Stevenson. RE; Davis. DP; Steiner. KG.
(2005). Physiological and foliar symptom response in the crowns of Primus serotina, Fraxinus americana
and Acer rubrum canopy trees to ambient ozone under forest conditions. Environ Pollut 133: 553-567.
http://dx.doi.0rg/10.1016/i.envpol.2004.06.012
Schraudner. M; Moeder. W: Wiese. C: Van Camp. W: Inze. D; Langebartels. C: Sandermann. H. Jr. (1998).
Ozone-induced oxidative burst in the ozone biomonitor plant, tobacco Bel W3. Plant J 16: 235-245.
http://dx.doi.org/10.1046/i. 1365-313x. 1998.00294.x
Severing. JF; Stich. K; Soja. G. (2007). Ozone stress and antioxidant substances in Trifolium repens and
Centaureajacea leaves. Environ Pollut 146: 707-714. http://dx.doi.Org/10.1016/j.envpol.2006.04.006
Sharkev. TD; Wiberlev. AE; Donohue. AR. (2008). Isoprene emission from plants: Why and how. Ann Bot
101: 5-18. http://dx.doi.org/10.1093/aob/mcm240
Singh. E; Tiwari. S: Agrawal. M. (2009). Effects of elevated ozone on photosynthesis and stomatal
conductance of two soybean varieties: A case study to assess impacts of one component of predicted global
climate change. Plant Biol (Stuttg) 11: 101-108. http://dx.doi.Org/10.llll/i.1438-8677.2009.00263.x
Singh. E; Tiwari. S: Agrawal. M. (2010a). Variability in antioxidant and metabolite levels, growth and yield of
two soybean varieties: An assessment of anticipated yield losses under projected elevation of ozone. Agric
Ecosyst Environ 135: 168-177. http://dx.doi.Org/10.1016/j.agee.2009.09.004
Sitch. S: Cox. PM; Collins. WJ: Huntingford. C. (2007). Indirect radiative forcing of climate change through
ozone effects on the land-carbon sink. Nature 448: 791-794. http://dx.doi.org/10.1038/nature06059
Skarbv. L; Ottosson. S: Karlsson. PE; Wallina. G: Sellden. G: Medina. EL; Pleijel. H. (2004). Growth of
Norway spruce (Picea abies) in relation to different ozone exposure indices: A synthesis. Atmos Environ
38: 2225-2236.
Skarbv. L; Troeng. E; Bostrom. CA. (1987). Ozone uptake and effects on transpiration, net photosynthesis, and
dark respiration in Scots pine. Forest Sci 33: 801-808.
Smith. G. (2012). Ambient ozone injury to forest plants in Northeast and North Central USA: 16 years of
biomonitoring. Environ Monit Assess 184: 4049-4065. http://dx.doi.org/10.1007/sl0661-011-2243-z
Smith. G: Coulston. J; Jepsen. E; Prichard. T. (2003). A national ozone biomonitoring program: Results from
field surveys of ozone sensitive plants in northeastern forests (1994-2000). Environ Monit Assess 87: 271-
291.
Soja. G: Barnes. JD; Posch. M; Vandermeiren. K; Pleijel. H; Mills. G. (2000). Phenological weighting of
ozone exposures in the calculation of critical levels for wheat, bean and plantain. Environ Pollut 109: 517-
524. http://dx.doi.org/10.1016/80269-7491(00)00055-5
Soja. G: Reichenauer. TG: Bid. M; Soja. AM; Schaber. R; Gangl. H. (2004). Long-term ozone exposure and
ozone uptake of grapevines in open-top chambers. Atmos Environ 38: 2313-2321.
Somers. GL; Chappelka. AH; Rosseau. P; Renfro. JR. (1998). Empirical evidence of growth decline related to
visible ozone injury. For Ecol Manage 104: 129-137. http://dx.doi.org/10.1016/S0378-1127(97)00252-l
Souza. L; Neufeld. HS: Chappelka. AH; Burkev. KO; Davison. AW. (2006). Seasonal development of ozone-
induced foliar injury on tall milkweed (Asclepias exaltata) in Great Smoky Mountains National Park.
Environ Pollut 141: 175-183. http://dx.doi.Org/10.1016/i.envpol.2005.07.022
Stampfli. A; Fuhrer. J. (2010). Spatial heterogeneity confounded ozone-exposure experiment in semi-natural
grassland. Oecologia 162: 515-522. http://dx.doi.org/10.1007/s00442-009-1462-2
Stewart. CA. (1998) Impact of ozone on the reproductive biology of Brassica campestris L and Plantago major
L. (Doctoral Dissertation). Loughborough University of Technology, England. Retrieved from
http://ethos.bl.uk/OrderDetails.do?did=l&uin=uk.bl.ethos.299673
9-181
-------
Stewart. CA; Black. VJ; Black. CR; Roberts. JA. (1996). Direct effects of ozone on the reproductive
development of Brassica species. J Plant Physiol 148: 172-178.
Stoelken. G: Pritsch. K; Simon. J; Mueller. CW: Grams. TEE; Esperschuetz. J; Gayler. S: Buegger. F;
Brueggemann. N; Meier. R; Zeller. B; Winkler. JB; Rennenberg. H. (2010). Enhanced ozone exposure of
European beech (Fagus sylvatica) stimulates nitrogen mobilization from leaf litter and nitrogen
accumulation in the soil. Plant Biosystems 144: 537-546. http://dx.doi.org/10.1080/11263500903429346
Street. NR; James. TM; James. T; Mikael. B; Jaakko. K; Mark. B; Taylor. G. (2011). The physiological,
transcriptional and genetic responses of an ozone-sensitive and an ozone tolerant poplar and selected
extremes of their F2 progeny. Environ Pollut 159: 45-54. http://dx.doi.Org/10.1016/i.envpol.2010.09.027
Summers. CG: Retzlaff. WA; Stephenson. S. (1994). The effect of ozone on the mean relative growth-rate of
Diuraphis-noxia (Mordvilko) (Homoptera, Aphididae). JAgrEntomol 11: 181-187.
Talhelm. AF; Pregitzer. KS: Zak. DR. (2009). Species-specific responses to atmospheric carbon dioxide and
tropospheric ozone mediate changes in soil carbon. Ecol Lett 12: 1219-1228.
htrp://dx.doi.org/10.1111/i.l461-0248.2009.01380.x
Tamaoki. M; Nakaiima. N; Kubo. A; Aono. M; Matsuvama. T; Saji. H. (2003). Transcriptome analysis of O3-
exposed Arabidopsis reveals that multiple signal pathways act mutually antagonistically to induce gene
expression. Plant Mol Biol 53: 443-456. http://dx.doi.Org/10.1023/B:PLAN.0000019064.55734.52
Temple. PJ: Kupper. RS: Lennox. RW: Rohr. K. (1988). Injury and yield responses of differentially irrigated
cotton to ozone. Agron J 80: 751-755. http://dx.doi.org/10.2134/agronjl988.00021962008000050011x
Temple. PJ: Riechers. GH; Miller. PR. (1992). Foliar injury responses of ponderosa pine seedlings to ozone,
wet and dry acidic deposition, and drought. Environ Exp Bot 32: 101-113. http://dx.doi.org/10.1016/0098-
8472(92)9003 5-Z
Theis. N; Raguso. RA. (2005). The effect of pollination on floral fragrance in thistles. J Chem Ecol 31: 2581-
2600.
Thomas. VFD; Braun. S: Fluckiger. W. (2005). Effects of simultaneous ozone exposure and nitrogen loads on
carbohydrate concentrations, biomass, and growth of young spruce trees (Picea abies). Environ Pollut 137:
507-516. http://dx.doi.0rg/10.1016/i.envpol.2005.02.002
Thomas. VFD; Braun. S: Fluckiger. W. (2006). Effects of simultaneous ozone exposure and nitrogen loads on
carbohydrate concentrations, biomass, growth, and nutrient concentrations of young beech trees (Fagus
sylvatica). Environ Pollut 143: 341-354. http://dx.doi.Org/10.1016/i.envpol.2005.ll.036
Tian. H; Melillo. J: Lu. C: Kicklighter. D; Liu. M; Ren. W: Xu. X; Chen. G: Zhang. C: Pan. S: Liu. J: Running.
S.. (2011). China's terrestrial carbon balance: Contributions from multiple global change factors. Global
Biogeochem Cycles 25: GB1007. httrj://dx.doi.org/10.1029/2010GB003838
Tingev. DT; Hogsett. WE; Lee. EH; Herstrom. AA; Azevedo. SH. (1991). An evaluation of various alternative
ambient ozone standards based on crop yield loss data. In RL Berglund; DR Lawson; DJ McKee (Eds.),
Tropospheric Ozone and the Environment (pp. 272-288). Pittsburgh, PA: Air & Waste Management
Association.
Tingev. DT; Hogsett. WE; Lee. EH; Laurence. JA. (2004). Stricter ozone ambient air quality standard has
beneficial effect on ponderosa pine in California. J Environ Manage 34: 397-405.
Tingev. DT: Johnson. MG: Lee. EH; Wise. C: Waschmann. R; Olszyk. DM; Watrud. LS: Donegan. KK.
(2006). Effects of elevated CO2 and O3 on soil respiration under ponderosa pine. Soil Biol Biochem 38:
1764-1778. http://dx.doi.org/10.1016/i.soilbio.2005.12.003
Tingev. DT: McVeetv. BD; Waschmann. R; Johnson. MG: Phillips. PL; Rygiewicz. PT; Olszyk. DM. (1996).
A versatile sun-lit controlled-environment facility for studying plant and soil processes. J Environ Qua! 25:
614-625.
Tingev. DT; Rodecap. KD; Lee. EH; Hogsett. WE; Gregg. JW. (2002). Pod development increases the ozone
sensitivity of Phaseolus vulgaris. Water Air Soil Pollut 139: 325-341.
9-182
-------
Tissue. DT; Griffin. KL; Ball. T. (1999). Photo synthetic adjustment in field-grown ponderosa pine trees after
six years of exposure to elevated CO2. Tree Physiol 19: 221-228.
Tissue. DT; Thomas. RB; Strain. BR. (1997). Atmospheric CO2 enrichment increases growth and
photosynthesis of Pinus taeda: A 4 year experiment in the field. Plant Cell Environ 20: 1123-1134.
http://dx.doi.0rg/10.1046/i.1365-3040.1997.d01-140.x
Tjoelker. MG: Volin. JC: Oleksyn. J; Reich. PB. (1995). Interaction of ozone pollution and light effects on
photosynthesis in a forest canopy experiment. Plant Cell Environ 18: 895-905.
http://dx.doi.org/10.1111/i. 1365-3040.1995.tb00598.x
Tobiessen. P. (1982). Dark opening of stomata in successional trees. Oecologia 52: 356-359.
http://dx.doi.org/10.1007/BF00367959
Toet. S: Ineson. P; Peacock. S: Ashmore. M. (2011). Elevated ozone reduces methane emissions from peatland
mesocosms. Global Change Biol 17: 288-296. http://dx.doi.Org/10.llll/i.1365-2486.2010.02267.x
Tong. D; Matfaur. R; Schere. K; Kang. D; Yu. S. (2007). The use of air quality forecasts to assess impacts of air
pollution on crops: Methodology and case study. Atmos Environ 41: 8772-8784.
http://dx.doi.0rg/10.1016/i.atmosenv.2007.07.060
Tong. DQ; Mauzerall. PL. (2008). Summertime state-level source-receptor relationships between nitrogen
oxides emissions and surface ozone concentrations over the continental United States. Environ Sci Technol
42: 7976-7984. http://dx.doi.org/10.1021/es7027636
Topa. MA; Vanderklein. DW: Corbin. A. (2001). Effects of elevated ozone and low light on diurnal and
seasonal carbon gain in sugar maple. Plant Cell Environ 24: 663-677.
Torsethaugen. G: Pell. EJ; Assmann. SM. (1999). Ozone inhibits guard cell K+ channels implicated in stomatal
opening. PNAS 96: 13577-13582.
Tosti. N; Pasqualini. S: Borgogni. A; Ederli. L; Falistocco. E; Crispi. S: Paolocci. F. (2006). Gene expression
profiles of O3-treated Arabidopsis plants. Plant Cell Environ 29: 1686-1702.
http://dx.doi.0rg/10.llll/i.1365-3040.2006.01542.x
Turnipseed. AA; Burns. SP; Moore. DJP; Hu. J; Guenther. AB; Monson. RK. (2009). Controls over ozone
deposition to a high elevation subalpine forest. Agr Forest Meteorol 149: 1447-1459.
http://dx.doi.0rg/10.1016/i.agrformet.2009.04.001
U.S. EPA (U.S. Environmental Protection Agency). (1977). Photochemical oxidant air pollutant effects on a
mixed conifer forest ecosystem: A progress report [EPA Report]. (EPA-600/3-77-104). Corvallis, OR.
http://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=20015IYPtxt
U.S. EPA (U.S. Environmental Protection Agency). (1978a). Air quality criteria for ozone and other
photochemical oxidants [EPA Report]. (EPA/600/8-78/004). Washington, DC.
U.S. EPA (U.S. Environmental Protection Agency). (1984). Air quality criteria for ozone and other
photochemical oxidants, volume III of V (review draft) [EPA Report]. (EPA-600/8-84-020A3). Research
Triangle Park, NC. http://www.ntis.gov/search/product. aspx? ABBR=PB85126050
U.S. EPA (U.S. Environmental Protection Agency). (1986). Air quality criteria for ozone and other
photochemical oxidants [EPA Report]. (EPA-600/8-84-020aF - EPA-600/8-84-020eF). Research Triangle
Park, NC. http://www.ntis.gov/search/product.aspx?ABBR=PB87142949
U.S. EPA (U.S. Environmental Protection Agency). (1996b). Air quality criteria for ozone and related
photochemical oxidants, Vol. II of III [EPA Report]. (EPA/600/P-93/004BF). Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (1996c). Air quality criteria for ozone and related
photochemical oxidants, Vol. Ill of III [EPA Report]. (EPA/600/P-93/004cF). Research Triangle Park, NC.
http://www.ntis.gov/search/product. aspx? ABBR=PB96185608
9-183
-------
U.S. EPA (U.S. Environmental Protection Agency). (1996e). Review of national ambient air quality standards
for ozone: Assessment of scientific and technical information: OAQPS staff paper [EPA Report].
(EPA/452/R-96/007). Research Triangle Park, NC.
http://www.ntis.gov/search/product. aspx?ABBR=PB9620343 5
U.S. EPA (U.S. Environmental Protection Agency). (2006b). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/R-05/004AF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
U.S. EPA (U.S. Environmental Protection Agency). (2007b). Review of the national ambient air quality
standards for ozone: Policy assessment of scientific and technical information: OAQPS staff paper [EPA
Report]. (EPA/452/R-07/003). Research Triangle Park, NC.
http://www.epa.gov/ttn/naaqs/standards/ozone/data/2007 01 ozone staff_paper
U.S. EPA (U.S. Environmental Protection Agency). (2008b). Integrated science assessment for oxides of
nitrogen and sulfur: Ecological criteria [EPA Report]. (EPA/600/R-08/082F). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=201485
U.S. Forest Service. National Park Service., and U.S. Fish and Wildlife Service.. (2000). Federal land
managers air quality related values work group (FLAG): Phase I report. Denver, CO: U.S. Forest Service.
http://www.nature.nps.gov/air/permits/flag/index.cfm
Uddling. J; Hogg. AJ; Teclaw. RM; Carroll. MA; Ellsworth. PS. (2010). Stomatal uptake of O3 in aspen and
aspen-birch forests under free-air CO2 and O3 enrichment. Environ Pollut 158: 2023-2031.
http://dx.doi.0rg/10.1016/i.envpol.2009.12.001
Uddling. J; Teclaw. RM; Kubiske. ME; Pregitzer. KS: Ellsworth. DS. (2008). Sap flux in pure aspen and mixed
aspen-birch forests exposed to elevated concentrations of carbon dioxide and ozone. Tree Physiol 28: 1231-
1243.
Uddling. J; Teclaw. RM: Pregitzer. KS: Ellsworth. DS. (2009). Leaf and canopy conductance in aspen and
aspen-birch forests under free-air enrichment of carbon dioxide and ozone. Tree Physiol 29: 1367-1380.
http://dx.doi.org/10.1093/treephvs/tpp070
UIUC (University of Illinois at Urbana-Champain). (2010). SoyFACE. Available online at
http://soyface.illinois.edu/ (accessed December 8, 2010).
UNECE (United Nations Economic Commission for Europe). (1988). ECE critical levels workshop; March;
Bad Harzburg, Germany. Geneva, Switzerland.
UNEP (United Nations Environment Programme). (2003). Ecosystems and human well-being: A framework
for assessment. Washington, DC: Island Press.
Unsworth. MH; Heagle. AS; Heck. WW. (1984a). Gas exchange in open-top field chambers: I. Measurement
and analysis of atmospheric resistances to gas exchange. Atmos Environ 18: 373-380.
http://dx.doi.org/10.1016/0004-6981(84)90111-2
Unsworth. MH; Heagle. AS; Heck. WW. (1984b). Gas exchange in open-top field chambers: II. Resistances to
ozone uptake by soybeans. Atmos Environ 18: 381-385. http://dx.doi.org/10.1016/0004-6981(84)90112-4
USDA (U.S. Department of Agriculture). (2011). Ozone biomonitoring program. Available online at
http://www.nrs.fs.fed.us/fia/topics/ozone/ (accessed January 28, 2011).
Vahisalu. T; Kollist. H; Wang. YF; Nishimura. N; Chan. WY; Valerio. G; Lamminmaki. A; Brosche. M;
Moldau. H; Desikan. R; Schroeder. JI; Kangasjarvi. J. (2008). SLAC1 is required for plant guard cell S-
type anion channel function in stomatal signalling. Nature 452: 487-491.
http://dx.doi.org/10.1038/nature06608
Valkama. E; Koricheva. J; Oksanen. E. (2007). Effects of elevated O3, alone and in combination with elevated
CO2, on tree leaf chemistry and insect herbivore performance: Ameta-analysis. Global Change Biol 13:
184-201. http://dx.doi.org/10.1111/i.l365-2486.01284.x
9-184
-------
van Buuren. ML; Guidi. L; Fornale. S: Ghetti. F; Franceschetti. M; Soldatini. GF; Bagni. N. (2002). Ozone-
response mechanisms in tobacco: Implications of polyamine metabolism. New Phytol 156: 389-398.
http://dx.doi.0rg/10.1046/i.1469-8137.2002.00539.x
Van Dingenen. R; Dentener. FJ; Raes. F; Krol. MC: Emberson. L; Cofala. J. (2009). The global impact of
ozone on agricultural crop yields under current and future air quality legislation. Atmos Environ 43: 604-
618. http://dx.doi.0rg/10.1016/j.atmosenv.2008.10.033
Vandermeiren. K; Black. C: Pleijel. H; de Temmerman. L. (2005). Impact of rising tropospheric ozone on
potato: Effects on photosynthesis, growth, productivity and yield quality. Plant Cell Environ 28: 982-996.
http://dx.doi.0rg/10.llll/i.1365-3040.2005.01316.x
Velikova. V: Pinelli. P; Pasqualini. S: Reale. L; Ferranti. F; Loreto. F. (2005). Isoprene decreases the
concentration of nitric oxide in leaves exposed to elevated ozone. New Phytol 166: 419-426.
http://dx.doi.0rg/10.llll/i.1469-8137.2005.01409.x
Vickers. CE; Possell. M; Cojocariu. CI; Velikova. VB; Laothawornkitkul. J; Ryan. A; Mullineaux. PM; Hewitt.
CN. (2009). Isoprene synthesis protects transgenic tobacco plants from oxidative stress. Plant Cell Environ
32: 520-531. http://dx.doi.Org/10.llll/i.1365-3040.2009.01946.x
Vigue. LM; Lindroth. RL. (2010). Effects of genotype, elevated CO2 and elevated O3 on aspen
phytochemistry and aspen leaf beetle Chrysomela crotchi performance. Agr Forest Entomol 12: 267-276.
http://dx.doi.0rg/10.llll/i.1461-9563.2010.00475.x
Volk. M; Bungener. P; Contat. F; Montani. M; Fuhrer. J. (2006). Grassland yield declined by a quarter in 5
years of free-air ozone fumigation. Global Change Biol 12: 74-83. http://dx.doi.org/10.1111/i. 1365-
2486.2005.01083.x
Volk. M; Geissmann. M; Blatter. A; Contat. F; Fuhrer. J. (2003). Design and performance of a free-air
exposure system to study long-term effects of ozone on grasslands. Atmos Environ 37: 1341-1350.
Volk. M; Obrist D; Novak. K; Giger. R; Bassin. S: Fuhrer. J. (2011). Subalpine grassland carbon dioxide
fluxes indicate substantial carbon losses under increased nitrogen deposition, but not at elevated ozone
concentration. Global Change Biol 17: 366-376. http://dx.doi.Org/10.llll/i.1365-2486.2010.02228.x
Vollenweider. P; Woodcock. H; Keltv. MJ; Hofer. RM. (2003). Reduction of stem growth and site dependency
of leaf injury in Massachusetts black cherries exhibiting ozone symptoms. Environ Pollut 125: 467-480.
Vollsnes. AV; Kruse. OMO; Eriksen. AB; Oxaal. U: Futsaether. CM. (2010). In vivo root growth dynamics of
ozone exposed Trifolium subterraneum. Environ Exp Bot 69: 183-188.
http://dx.doi.0rg/10.1016/i.envexpbot.2010.03.007
von Tiedemann. A. (1996). Single and combined effects of nitrogen fertilization and ozone on fungal leaf
diseases on wheat. Z Pflanzenkrankh Pflanzenschutz 103: 409-419.
Vuorinen. T; Nerg. AM; Holopainen. JK. (2004). Ozone exposure triggers the emission of herbivore-induced
plant volatiles, but does not disturb tritrophic signalling. Environ Pollut 131: 305-311.
http://dx.doi.0rg/10.1016/i.envpol.2004.02.027
Wallin. G: Skarbv. L. (1992). The influence of ozone on the stomatal and non-stomatal limitation of
photosynthesis in Norway spruce, Picea abies (L.) Karst, exposed to soil moisture deficit. Trees Struct
Funct 6: 128-136. http://dx.doi.org/10.1007/BF00202428
Wang. L; He. X; Chen. W. (2009b). Effects of elevated ozone on photosynthetic CO2 exchange and
chlorophyll a fluorescence in leaves of Quercus mongolica grown in urban area. Bull Environ Contam
Toxicol 82: 478-481. http://dx.doi.org/10.1007/s00128-008-9606-3
Wang. X; Mauzerall. PL. (2004). Characterizing distributions of surface ozone and its impact on grain
production in China, Japan and South Korea: 1990 and 2020. Atmos Environ 38: 4383-4402.
http://dx.doi.0rg/10.1016/i.atmosenv.2004.03.067
Wang. X; Taub. DR. (2010). Interactive effects of elevated carbon dioxide and environmental stresses on root
mass fraction in plants: Ameta-analytical synthesis using pairwise techniques. Oecologia 163: 1-11.
http://dx.doi.org/10.1007/s00442-010-1572-x
9-185
-------
Wang. X; Zheng. Q; Feng. Z; Xie. J: Feng. Z; Ouyang. Z; Manning. WJ. (2008). Comparison of a diurnal vs
steady-state ozone exposure profile on growth and yield of oilseed rape (Brassica napus L.) in open-top
chambers in the Yangtze Delta, China. Environ Pollut 156: 449-453.
http://dx.doi.0rg/10.1016/i.envpol.2008.01.027
Watanabe. M; Yamaguchi. M; Tabe. C: Iwasaki. M; Yamashita. R; Funada. R; Fukami. M; Matsumura. H;
Kohno. Y; Izuta. T. (2007). Influences of nitrogen load on the growth and photo synthetic responses of
Quercus serrata seedlings to O3. Trees Struct Funct 21: 421-432. http://dx.doi.org/10.1007/s00468-007-
0134-2
Weinstein. DA; Gollands. B; Retzlaff. WA. (2001). The effects of ozone on a lower slope forest of the Great
Smoky Mountain National Park: Simulations linking an individual tree model to a stand model. Forest Sci
47: 29-42.
Weinstein. DA; Laurence. JA; Retzlaff. WA; Kern. JS: Lee. EH; Hogsett. WE; Weber. J. (2005). Predicting the
effects of tropospheric ozone on regional productivity of ponderosa pine and white fir. For Ecol Manage
205: 73-89. http://dx.doi.Org/10.1016/j.foreco.2004.10.007
Werner. H; Fabian. P. (2002). Free-air fumigation of mature trees: A novel system for controlled ozone
enrichment in grown-up beech and spruce canopies. Environ Sci Pollut Res Int 9: 117-121.
Wesely. ML; Hicks. BB. (2000). A review of the current status of knowledge on dry deposition [Review].
Atmos Environ 34: 2261-2282. http://dx.doi.org/10.1016/81352-2310(99)00467-7
Whitfield. CP; Davison. AW: Ashenden. TW. (1996). Interactive effects of ozone and soil volume on Plantago
major. New Phytol 134: 287-294. http://dx.doi.Org/10.llll/i.1469-8137.1996.tb04633.x
Whitfield. CP: Davison. AW: Ashenden. TW. (1997). Artificial selection and heritability of ozone resistance in
two populations of Plantago major. New Phytol 137: 645-655.
Wieser. G: Havranek. WM. (1995). Environmental control of ozone uptake in Larix decidua Mill:
A comparison between different altitudes. Tree Physiol 15: 253-258.
Wieser. G: Manning. WJ: Tausz. M; Bytnerowicz. A. (2006). Evidence for potential impacts of ozone on Pinus
cembra L. at mountain sites in Europe: An overview. Environ Pollut 139: 53-58.
http://dx.doi.0rg/10.1016/i.envpol.2005.04.037
Wilkinson. S: Davies. WJ. (2010). Drought, ozone, ABA and ethylene: New insights from cell to plant to
community [Review]. Plant Cell Environ 33: 510-525. http://dx.doi.Org/10.llll/i.1365-3040.2009.02052.x
Will. RE; Ceulemans. R. (1997). Effects of elevated CO2 concentration on photosynthesis, respiration and
carbohydrate status of coppice Populus hybrids. Physiol Plant 100: 933-939.
http://dx.doi.org/10.1111/i. 1399-3054.1997.tb00020.x
Winner. WE; Lefohn. AS: Cotter. IS: Greitner. CS: Nellessen. J: McEvoy. LR. Jr; Olson. RL; Atkinson. CJ:
Moore. LD. (1989). Plant responses to elevational gradients of O3 exposures in Virginia. PNAS 86: 8828-
8832.
Wittig. VE; Ainsworth. EA; Long. SP (2007). To what extent do current and projected increases in surface
ozone affect photosynthesis and stomatal conductance of trees? A meta-analytic review of the last 3
decades of experiments [Review]. Plant Cell Environ 30: 1150-1162. http://dx.doi.org/10.1111/i. 1365-
3040.2007.01717.x
Wittig. VE; Ainsworth. EA; Naidu. SL; Karnosky. DF; Long. SP. (2009). Quantifying the impact of current and
future tropospheric ozone on tree biomass, growth, physiology and biochemistry: A quantitative meta-
analysis. Global Change Biol 15: 396-424. http://dx.doi.Org/10.llll/i.1365-2486.2008.01774.x
Woo. SY; Hincklev. TM. (2005). The effects of ozone on growth and stomatal response in the F-2 generation
of hybrid poplar (Populus trichocarpa x Populus deltoides). Biol Plantarum 49: 395-404.
http://dx.doi.org/10.1007/sl0535-005-0014-9
Wright. GA; Lutmerding. A; Dudareva. N; Smith. BH. (2005). Intensity and the ratios of compounds in the
scent of snapdragon flowers affect scent discrimination by honeybees (Apis mellifera). J Comp Physiol
ANeuroethol Sens Neural Behav Physiol 191: 105-114.
9-186
-------
Yamaguchi. M; Watanabe. M; Iwasaki. M; Tabe. C: Matsumura. H; Kohno. Y; Izuta. T. (2007). Growth and
photosynthetic responses of Fagus crenata seedlings to O3 under different nitrogen loads. Trees Struct
Funct 21: 707-718. http://dx.doi.org/10.1007/s00468-007-0163-x
Yan. K; Chen. W: He. XY; Zhang. GY; Xu. S: Wang. LL. (2010). Responses of photosynthesis, lipid
peroxidation and antioxidant system in leaves of Quercus mongolicato elevated O3. Environ Exp Bot 69:
198-204. http://dx.doi.0rg/10.1016/i.envexpbot.2010.03.008
Yoshida. S: Tamaoki. M; loki. M; Ogawa. D; Sato. Y; Aono. M; Kubo. A; Saii. S: Saji. H; Satoh. S: Nakajima.
N.. (2009). Ethylene and salicylic acid control glutathione biosynthesis in ozone-exposed Arabidopsis
thaliana. Physiol Plant 136: 284-298. http://dx.doi.Org/10.llll/i.1399-3054.2009.01220.x
Younglove. T; McCool. PM; Musselman. RC: Kahl. ME. (1994). Growth-stage dependent crop yield response
to ozone exposure. Environ Pollut 86: 287-295. http://dx.doi.org/10.1016/0269-7491(94)90169-4
Yuan. JS: Himanen. SJ: Holopainen. JK; Chen. F; Stewart. CN. Jr. (2009). Smelling global climate change:
Mitigation of function for plant volatile organic compounds [Review]. Trends Ecol Evol 24: 323-331.
http://dx.doi.0rg/10.1016/i.tree.2009.01.012
Yun. SC: Laurence. JA. (1999). The response of sensitive and tolerant clones of Populus tremuloides to
dynamic ozone exposure under controlled environmental conditions. New Phytol 143: 305-313.
Zak. PR; Holmes. WE; Pregitzer. KS. (2007). Atmospheric CO2 and O3 alter the flow of N15 in developing
forest ecosystems. Ecology 88: 2630-2639.
Zhang. C: Tian. HQ; Chappelka. AH; Ren. W: Chen. H; Pan. SF; Liu. ML; Stvers. DM; Chen. GS: Wang. YH.
(2007a). Impacts of climatic and atmospheric changes on carbon dynamics in the Great Smoky Mountains
National Park. Environ Pollut 149: 336-347. http://dx.doi.Org/10.1016/j.envpol.2007.05.028
Zhang. J; Schaub. M; Ferdinand. JA; Skelly. JM; Steiner. KG: Savage. JE. (2010a). Leaf age affects the
responses of foliar injury and gas exchange to tropospheric ozone in Primus serotina seedlings. Environ
Pollut 158: 2627-2634. http://dx.doi.Org/10.1016/i.envpol.2010.05.003
Zheng. F; Wang. X; Lu. F; Hou. P; Zhang. W: Duan. X; Zhou. X; Ai. Y; Zheng. H; Ouvang. Z; Feng. Z.
(2011). Effects of elevated ozone concentration on methane emission from a rice paddy in Yangtze River
Delta, China. Global Change Biol 17: 898-910. http://dx.doi.Org/10.llll/i.1365-2486.2010.02258.x
9-187
-------
10 THE ROLE OF TROPOSPHERIC OZONE IN CLIMATE
CHANGE AND UV-B SHIELDING EFFECTS
10.1 Introduction
Atmospheric O3 plays an important role in the Earth's energy budget by interacting
with incoming solar radiation and outgoing infrared radiation. Over mid-latitudes,
approximately 90% of the total atmospheric O3 column is located in the stratosphere
(Kar et al.. 2010: Crist et al.. 1994). Therefore, tropospheric O3 makes up a relatively
small portion (-10%) of the total column of O3 over mid-latitudes, but it does play
an important role in the overall radiation budget. The next section (Section 10.2)
briefly describes the physics of the earth's radiation budget, providing background
material for the subsequent two sections assessing how perturbations in tropospheric
O3 concentrations might affect (1) climate through its role as a greenhouse gas
(Section 10.3). and (2) health, ecology and welfare through its role in shielding the
earth's surface from solar ultraviolet radiation (Section 10.4). The concluding section
in this chapter (Section 10.5) includes a summary of effects assessed in this chapter
along with their associated causal determinations.
10.2 Physics of the Earth's Radiation Budget
Radiant energy from the sun enters the atmosphere in a range of wavelengths, but
peaks strongly in the visible (400-750 nm) part of the spectrum. Longer wavelength
infrared (750 nm-1 mm) and shorter wavelength ultraviolet (100-400 nm) radiation
are also present in the solar electromagnetic spectrum. Since the energy possessed by
a photon is inversely proportional to its wavelength, infrared (IR) radiation carries
the least energy per photon, and ultraviolet (UV) radiation carries the most energy
per photon. Ultraviolet radiation is further subdivided into classes (bands) based on
wavelength: UV-A refers to wavelengths from 400-315 nm; UV-B from 315-
280 nm; and UV-C from 280-100 nm. Within the UV spectrum, UV-A radiation is
the least energetic band and UV-C is the most energetic band.
The wavelength of radiation also determines how the photons interact with the
complex mixture of gases, clouds, and particles present in the atmosphere (see
Figure 10-1). UV-A radiation can be scattered but is not absorbed to any meaningful
degree by atmospheric gases including O3. UV-B radiation is absorbed and scattered
in part within the atmosphere. UV-C is almost entirely blocked by the Earth's upper
atmosphere, where it participates in photoionization and photodissociation processes
including absorption by stratospheric O3. Since UV-A radiation is less energetic and
does not interact with O3 in the troposphere or the stratosphere and UV-C radiation is
almost entirely blocked by stratospheric O3, UV-B radiation is the most important
band to consider in relation to tropospheric O3 shielding.
10-1
-------
Tropospheric O3 plays a "disproportionate" role in absorbing UV-B radiation
compared with stratospheric O3 on a molecule per molecule basis (Balis et al., 2002;
Zerefos et al.. 2002: Crist etal.. 1994: Bruhl and Crutzen. 1989). This effect results
from the higher atmospheric pressure present in the troposphere, resulting in higher
concentrations of gas molecules present that can absorb or scatter radiation. For this
reason, the troposphere is referred to as a "multiple scattering" regime for UV
absorption, compared to the "single scattering" regime in the stratosphere. Thus,
careful quantification of atmospheric absorbers and scatterers, along with a well-
resolved description of the physics of these interactions, is necessary for predicting
the effects of tropospheric O3 on UV-B flux at the surface.
Solar flux at all wavelengths has a temporal dependence, while radiative scattering
and absorption have strong wavelength, path length, and gas/particle concentration
dependencies. These combine to create nonlinear effects on UV flux at the Earth's
surface. Chapter 10 of the 2006 O3 AQCD (U.S. EPA. 2006b) describes in detail
several key factors that influence the spatiotemporal distribution of ground-level UV
radiation flux, including: (1) long-term solar activity including sunspot cycle;
(2) solar rotation; (3) the position of the Earth in its orbit around the sun;
(4) atmospheric absorption and scattering of UV radiation by gas molecules and
aerosol particles; (5) absorption and scattering by stratospheric and tropospheric
clouds; and (6) surface albedo. The efficiencies of absorption and scattering are
highly dependent on the concentration of the scattering medium, particle size (for
aerosols and clouds), and the altitude at which these processes are occurring. These
properties are sensitive to meteorology, which introduces additional elements of
spatial and temporal dependency in ground-level UV radiation flux.
10-2
-------
Vl
Backscattered
Radiation
Incident Solar UV Radiation
Stratospheric O3
Source: 2006 O3 AQCD (U.S. EPA. 2006b).
Figure 10-1 Diagram of the factors that determine human exposure to
ultraviolet radiation.
About 30% of incoming solar radiation is directly reflected back to space, mainly by
clouds or surfaces with high albedo (reflectivity), such as snow, ice, and desert sand.
Radiation that does penetrate to the Earth's surface and is absorbed can be re-emitted
in the longwave (infrared) portion of the spectrum; the rest goes into evaporating
water or soil moisture or emerges as sensible heat. The troposphere is opaque to the
outgoing longwave radiation. Polyatomic gases such as water vapor, CO2, CH4, and
O3 absorb and re-emit the radiation upwelling from the Earth's surface, reducing the
efficiency with which that energy returns to space. In effect, these gases act as a
blanket warming the Earth's surface. This phenomenon, known as the "Greenhouse
Effect," was first quantified in the 19th century (Arrhenius, 1896), and gives rise to
the term "greenhouse gas." The most important greenhouse gas is water vapor.
10-3
-------
10.3 Effects of Tropospheric O3 on Climate
10.3.1 Background
As a result of its interaction with incoming solar radiation and outgoing longwave
radiation, tropospheric O3 plays a major role in determining climate, and increases in
its abundance may contribute to climate change (IPCC. 2007c). Models estimate that
the global average concentration of O3 in the troposphere has increased 30-70%
since the pre-industrial era (Gauss et al.. 2006). while observations indicate that in
some regions tropospheric O3 concentrations may have increased by factors as great
as 4 or 5 (Marenco et al.. 1994: Staehelin et al.. 1994). These increases are tied to the
rise in emissions of O3 precursors from human activity, mainly fossil fuel
consumption and agricultural processes.
The effect on climate of the tropospheric O3 concentration change since
pre-industrial times has been estimated to be about 25-40% of the anthropogenic
CO2 effect and about 75% of the anthropogenic CH4 effect OPCC. 2007c). ranking it
third in importance behind these two major greenhouse gases. In the 21st century, as
the Earth's population continues to grow and energy technology spreads to
developing countries, a further rise in the global concentration of tropospheric O3 is
likely, with associated consequences for human health and ecosystems relating to
climate change.
To examine the science of a changing climate and to provide balanced and rigorous
information to policy makers, the World Meteorological Organization (WMO) and
the United Nations Environment Programme (UNEP) formed the Intergovernmental
Panel on Climate Change (IPCC) in 1988. The IPCC supports the work of the
Conference of Parties (COP) to the United Nations Framework Convention on
Climate Change (UNFCCC). The IPCC periodically brings together climate
scientists from member countries of WMO and the United Nations to review
knowledge of the physical climate system, past and future climate change, and
evidence of human-induced climate change. IPCC climate assessment reports are
issued every five to seven years.
This section draws in part on the fourth IPCC Assessment Report (AR4) (IPCC.
2007c). as well as other peer-reviewed published research. Section 10.3.2 reviews
evidence of climate change in the recent past and projections of future climate
change. It also offers a brief comparison of tropospheric O3 relative to other
greenhouse gases. Section 10.3.3 describes factors that influence the magnitude of
tropospheric O3 effects on climate. Section 10.3.4 considers the competing effects of
O3 precursors on climate. Finally, Section 10.3.5 and Section 10.3.6 describe the
effects of changing tropospheric O3 concentrations on past and future climate.
Downstream effects resulting from climate change, such as ecosystem responses, are
outside the scope of this assessment, which focuses rather on the effects of changes
in tropospheric O3 concentrations on radiative forcing and climate.
10-4
-------
10.3.2 Climate Change Evidence and the Influence of Tropospheric O3
10.3.2.1 Climate Change in the Recent Past
From the end of the last ice age 12,000 years ago until the mid-1800s, observations
from ice cores show that concentrations of the long-lived greenhouse gases CO2,
CH4, and N2O have been relatively stable. Unlike these greenhouse gases, O3 is not
preserved in ice, and no record of it before the late 1800s exists. Models, however,
suggest that it, too, has remained relatively constant during this time period
(Thompson et al., 1993; Thompson, 1992). The stable mix of these greenhouse gases
in the atmosphere, together with water vapor, has kept the global mean temperature
of the Earth close to 15°C. Without the presence of greenhouse gases in the
atmosphere, the Earth's global mean temperature would be about 30°C cooler, or -
15°C.
Since the start of the Industrial Revolution, human activity has led to observable
increases of greenhouse gases in the atmosphere, mainly through fossil fuel
combustion. According to the IPCC AR4 (IPCC, 2007c), there is now "very high
confidence" that the net effect of anthropogenic greenhouse gas emissions since 1750
has led to warming, and it is "very likely" that human activity contributed to the
0.76°C rise in global mean temperature observed over the last century. The increase
of tropospheric O3 abundance may have contributed 0.1-0.3°C warming to the global
climate during this time period (Hansen et al., 2005; Mickley et al., 2004). Global
cooling due to anthropogenic aerosols (IPCC, 2007c) has likely masked the full
warming effect of the anthropogenic greenhouse gases on a global scale.
10.3.2.2 Projections of Future Climate Change
The IPCC AR4 projects a warming of ~0.2°C per decade for the remainder of the
21st century (IPCC. 2007c). Even at constant concentrations of greenhouse gases in
the atmosphere, temperatures are expected to increase by about 0.1°C per decade, due
to the slow response of oceans to the warming applied so far. It is likely that the
Earth will experience longer and more frequent heat waves in the 21st century,
together with more frequent droughts and/or heavy precipitation events in some
regions, due to perturbations in the hydrological cycle that result from changing
temperatures. Sea levels could increase by 0.3-0.8 meters by 2300 due to thermal
expansion of the oceans. The extent of Arctic sea ice is expected to decline, and
contraction of the Greenland ice sheet could further contribute to the sea level rise
(IPCC. 2007c).
Projections of future climate change are all associated with some degree of
uncertainty. A major uncertainty involves future trends in the anthropogenic
emissions of greenhouse gases or their precursors. For the IPCC AR4 climate
projections, a set of distinct "storylines" or emission pathways was developed (IPCC.
10-5
-------
2000). Each storyline took into account factors such as population growth, mix of
energy technologies, and the sharing of technology between developed and
developing nations, and each resulted in a different scenario for anthropogenic
emissions. When these trends in emissions are applied to models, these scenarios
yield a broad range of possible climate trajectories for the 21st century.
A second factor bringing large uncertainty to model projections of future climate is
the representation of climate and, especially, climate feedbacks. A rise in surface
temperatures would perturb a suite of other processes in the earth-atmosphere-ocean
system, which may in turn either amplify the temperature increase (positive
feedback) or diminish it (negative feedback). One important feedback involves the
increase of water vapor content of the atmosphere that would accompany higher
temperatures (Bony et al., 2006). Water vapor is a potent greenhouse gas; accounting
for the water vapor feedback may increase the climate sensitivity to a doubling of
CO2 by nearly a factor of two (Held and Soden, 2000). The ice-albedo feedback is
also strongly positive; a decline in snow cover and sea ice extent would diminish the
Earth's albedo, allowing more solar energy to be retained at the surface (Holland and
Bitz, 2003; Rind et al., 1995). A final example of a climate feedback involves the
effects of changing cloud cover in a warming atmosphere. Models disagree on the
magnitude and even the sign of the cloud cover feedback on surface temperatures
(Soden and Held 2006).
10.3.2.3 Metrics of Potential Climate Change
Two metrics frequently used to estimate the potential climate effect of some
perturbation such as a change in greenhouse gas concentration are: (1) radiative
forcing; and (2) global warming potential (GWP). These metrics differ in a
fundamental way as described below.
Radiative forcing is a change in the radiative balance at a particular level of the
atmosphere or at the surface when a perturbation is introduced in the earth-
atmosphere-ocean system. In the global mean, radiative forcing of greenhouse gases
at the tropopause (top of the troposphere) is roughly proportional to the surface
temperature response (Hansen et al.. 2005: NRC. 2005). It thus provides a useful
metric for policymakers for assessing the response of the earth's surface temperature
to a given change in the concentration of a greenhouse gas. Positive values of
radiative forcing indicate warming in a test case relative to the control; negative
values indicate cooling. The units of radiative forcing are energy flux per area, or
W/m2.
Radiative forcing requires just a few model years to calculate, and it shows
consistency from model to model. However, radiative forcing does not take into
account the climate feedbacks that could amplify or dampen the actual surface
temperature response, depending on region. Quantifying the change in surface
temperature requires a climate simulation in which all important feedbacks are
accounted for. As some of these processes are not well understood, the surface
10-6
-------
temperature response to a given radiative forcing can be highly uncertain and can
vary greatly among models and even from region to region within the same model.
GWP indicates the integrated radiative forcing over a specified period (usually
100 years) from a unit mass pulse emission of a greenhouse gas or its precursor, and
is reported as the magnitude of this radiative forcing relative to that of CO2. GWP is
most useful for comparing the potential climate effects of long-lived gases, such as
N2O or CH4. Since tropospheric O3 has a lifetime on the order of weeks to months,
GWP is not seen as a valuable metric for quantifying the importance of O3 on
climate (Forster et al., 2007). Thus, this assessment focuses on radiative forcing as
the metric of climate influence resulting from changes in tropospheric O3
concentrations.
10.3.2.4 Tropospheric O3 as a Greenhouse Gas
Tropospheric O3 differs in important ways from other greenhouse gases. It is not
emitted directly, but is produced through photochemical oxidation of CO, CH4, and
nonmethane volatile organic compounds (VOCs) in the presence of nitrogen oxide
radicals (NOX = NO + NO2; see Chapter 3., Section 3.2 for further details on the
chemistry of O3 formation). It is also supplied by vertical transport from the
stratosphere. The lifetime of O3 in the troposphere is typically a few weeks, resulting
in an inhomogeneous distribution that varies seasonally; the distribution of the long-
lived greenhouse gases like CO2 and CH4 are much more uniform. The longwave
radiative forcing by O3 is mainly due to absorption in the 9.6 um window, where
absorption by water vapor is weak. It is therefore less sensitive to local humidity than
the radiative forcing by CO2 or CH4, for which there is much more overlap with the
water absorption bands (Lenoble, 1993). And unlike other major greenhouse gases,
O3 absorbs in the shortwave as well as the longwave part of the spectrum.
Figure 10-2 shows the main steps involved in the influence of tropospheric O3 on
climate. Emissions of O3 precursors including CO, VOCs, CH4, and NOX lead to
production of tropospheric O3. A change in the abundance of tropospheric O3
perturbs the radiative balance of the atmosphere, an effect quantified by the radiative
forcing metric. The earth-atmosphere-ocean system responds to the radiative forcing
with a climate response, typically expressed as a change in surface temperature.
Finally, the climate response causes downstream climate-related health and
ecosystem effects, such as redistribution of diseases or ecosystem characteristics due
to temperature changes. Feedbacks from both the climate response and downstream
effects can, in turn, affect the abundance of tropospheric O3 and O3 precursors
through multiple mechanisms. Direct feedbacks are discussed further in
Section 10.3.3.4: the downstream climate effects and their long-term feedbacks are
extremely complex and outside the scope of this assessment.
10-7
-------
Precursor Emissions of
CO, VOCs, CH4, NOX
(Tg/y)
Changes inTropospheric
O, Abundance
(Tg)
Due to O3 Change
Radiative Forcing
oO3Ch
(W/m2)
^^
t
Climate Response
^^
^^
«i:;: i ,.-;-., ir-
Note: This figure includes the relationship between precursor emissions, changes in tropospheric O3 abundance, radiative forcing,
climate response, and climate effects. Units shown are those typical for each quantity illustrated. Feedbacks from both the climate
response and climate effects can, in turn, affect the abundance of tropospheric O3 and O3 precursors through multiple feedback
mechanisms. Climate effects and their feedbacks are deemphasized in the figure since these downstream effects are extremely
complex and outside the scope of this assessment.
Figure 10-2 Schematic illustrating the effects of tropospheric O3 on climate.
The IPCC (2007c) reported a radiative forcing of 0.35 W/m2 for the change in
tropospheric O3 abundance since the pre-industrial era, ranking it third in importance
after the greenhouse gases CO2 (1.66 W/m2) and CH4 (0.48 W/m2). Figure 10-3
shows the global average radiative forcing estimates and uncertainty ranges in 2005
for anthropogenic CO2, CH4, O3 and other important agents and mechanisms.
The error bars encompassing the tropospheric O3 radiative forcing estimate in the
figure range from 0.25 to 0.65 W/m2, making it relatively more uncertain than the
long-lived greenhouse gases.
10-8
-------
RADIATIVE FORCING COMPONENTS
RF Terms
Long-lived
greenhouse gases
Ozone
Stratospheric water
vapour trom CH,,
Surface albedo
< Direct ettect
Total
Aerosol
Cloud albedo
effect
Linear contrails
Solar irradianco
Total nci
anthropogenic
RF values (W rrT) Spatial scale LOSU
1.66 [1.49 to 1.83]
0.48 [0.43 to 0.53]
0.16 [0.14 to 0.18]
-0.05 [-0.15 to 0.05]
0.35 [0.25 to 0.65]
0.07 [0.02 to 0.12]
-0.2 [-0.4 to 0.0]
0.1 [0.0 to 0.2]
-OS [-0.9 to-0.1]
-0.7 [-1.8 to-0.3]
0.01 [0.003 to 0.03]
0.12 [0.06 to 0.30]
1.6 [0.6 to 2.4]
Global
Global
Continental
to global
Global
Local to
continental
Continental
to global
Continental
to global
Continental
Global
High
High
Med
- Low
Med
-Low
Low
-2-1012
Radiative Forcing (W rrr2)
Note: This figure shows the typical geographical extent (spatial scale) of the radiative forcing and the assessed level of scientific
understanding (LOSU). The net anthropogenic radiative forcing and its range are also shown. These require summing asymmetric
uncertainty estimates from the component terms, and cannot be obtained by simple addition. Additional radiative forcing factors
not included here are considered to have a very low LOSU.
Source: Reprinted with permission of Cambridge University Press (IPCC. 2007c).
Figure 10-3 Global average radiative forcing (RF) estimates and uncertainty
ranges in 2005 for anthropogenic 062, CH4, O$, and other
important agents and mechanisms.
10-9
-------
10.3.3 Factors that Influence the Effect of Tropospheric O$ on Climate
This section describes the main factors that influence the magnitude of the climate
response to changes in tropospheric O3 abundance. They include: (1) trends in the
concentration of tropospheric O3; (2) the effect of surface albedo on O3 radiative
forcing; (3) the effect of vertical distribution on O3 radiative forcing; (4) feedback
factors that can alter the climate response to O3 radiative forcing; and (5) the indirect
effects of tropospheric O3 on the carbon cycle. Trends in stratospheric O3 abundance
may also affect temperatures at the Earth's surface, but aside from issues relating to
stratospheric-tropospheric exchange discussed in Chapter 3_, Section 3.4.1.1,
stratospheric O3 assessment is beyond the scope of this document.
10.3.3.1 Trends in the Concentration of Tropospheric Os
To first order, the effect of tropospheric O3 on global surface temperature is
proportional to the change in tropospheric O3 concentration. The earth's surface
temperatures are most sensitive to O3 abundance perturbations in the mid to upper
troposphere. This section therefore focuses mainly on observed O3 concentration
trends in the free troposphere or in regions far from O3 sources, where a change in
O3 concentrations may indicate change throughout the troposphere. Data from
ozonesondes, mountaintops, and remote surface sites are discussed, as well as
satellite data.
Observed Trends in O3 Concentrations since the Pre-lndustrial Era
Measurements of O3 concentrations at two European mountain sites dating from the
late 1800s to early 1900s show values at about 10 ppb, about one-fifth the values
observed today at similar sites (Pavelin et al.. 1999: Marenco et al.. 1994).
The accuracy of these early measurements is questionable however, in part because
they exhibit O3 concentrations equivalent to or only a couple of parts per billion
greater than those observed at nearby low-altitude sites during the same time period
(Mickley et al.. 2001: Volz and Kiev. 1988). A larger vertical gradient in
tropospheric O3 concentration would be expected because of its stratospheric source
and its longer lifetime aloft. In another study, Staehelin et al. (1994) revisited
observations made in the Swiss mountains during the 1950s and found a doubling in
O3 concentrations from that era to 1989-1991.
Routine observations of O3 in the troposphere began in the 1970s with the use of
balloon-borne ozonesondes, but even this record is sparse. Trends from ozonesondes
have been highly variable and dependent on region (Logan et al., 1999). Over most
sites in the U.S., ozonesondes reveal little trend. Over Canada, observations show a
decline in O3 concentrations between 1980 and 1990, then a rebound in the following
decade (Tarasick et al.. 2005). Ozonesondes over Europe give a mixed picture.
European ozonesondes showed increases in the 1970s and 1980s, with smaller
10-10
-------
increases or even declines since then (Oltmans et al., 2006; Logan et al., 1999). Over
Japan, O3 concentrations in the lower troposphere increased about 0.2-0.4 ppb/year
during the 1990s (Naja and Akimoto, 2004).
Ground-based measurements in remote regions provide a record of tropospheric O3
concentrations, but like ozonesonde data, such measurements are sparse before the
1970s. Springtime O3 observations from several mountain sites in the western U.S.
show a positive trend of about of 0.5-0.7 ppb/year since the 1980s (Cooper et al.,
2010; Jaffe et al., 2003). Ship-borne O3 measurements for the time period 1977 to
2002 indicate increases of 0.1-0.7 ppb/year over much of the Atlantic south of 40°N,
but no appreciable change north of 40°N (Lelieveld et al., 2004). The lack of trend
for the North Atlantic would seem at odds with O3 observations at Mace Head
(53°N) on the west coast of Ireland, which show a significant positive trend of about
0.5 ppb/year from 1987 to 2003 (Simmonds et al.. 2004). Over Japan, O3
concentrations at a remote mountain site have increased 1 ppb/year from 1998 to
2003 (Tanimoto, 2009), a rate more than double that recorded by ozonesondes in the
lower troposphere over Japan during the 1990s (Naja and Akimoto, 2004).
At Zugspitze, a mountain site in Germany, O3 concentrations increased by 12% per
decade during the 1970s and 1980s, consistent with European ozonesondes (Oltmans
et al., 2006). Since then, O3 concentrations continue to increase at Zugspitze, but
more slowly. What little data exist for the Southern Hemisphere point to measurable
increases in tropospheric O3 concentrations in recent decades, as much as -15% at
Cape Grim in the 1989-2004 time period (Oltmans et al., 2006).
The satellite record is now approaching a length that can be useful for diagnosing
trends in the total tropospheric O3 column (details on the use of satellites to measure
tropospheric O3 concentrations are covered in Chapter 3_, Section 3.5.5.5). In contrast
to the surface data from ships, tropospheric O3 columns from the Total Ozone
Mapping Spectrometer (TOMS) show no trend over the tropical Atlantic for the
period 1980-1990 (Thompson and Hudson, 1999). Over the Pacific Ocean, a longer,
25 year record of TOMS data again reveals no trend over the tropics, but shows
increases in tropospheric column O3 of about 2-3 Dobson Units (DU)1 at mid-
latitudes in both hemispheres (Ziemke et al., 2005).
Interpreting these recent trends in tropospheric O3 concentrations is challenging.
The first difficulty is reconciling apparently contradictory trends in the observations,
e.g., over tropical oceans. A second difficulty is that the O3 concentration trends
depend on several factors, not all of which can be well characterized. These factors
include (1) trends in emissions of O3 precursors, (2) variation in the stratospheric
source of O3, (3) changes in solar radiation resulting from stratospheric O3 depletion,
and (4) trends in tropospheric temperatures (Fusco and Logan. 2003). Recent positive
trends in the western U.S. and over Japan are consistent with the rapid increase in
emissions of O3 precursors from mainland Asia and transport of pollution across the
1 The Dobson Unit is a typical unit of measure for the total O3 in a vertical column above the Earth's surface. One DU is equivalent to
the amount of O3 that would exist in a 1 urn (1CT5 meter) thick layer of pure O3 at standard temperature (0°C) and pressure (1
atm), and corresponds to a column of O3 containing 2.69 x 1 o20 molecules/m2. A typical value for the amount of O3 in a column of
the Earth's atmosphere, although highly variable, is 300 DU and approximately 10% (30 DU) of that exists in the troposphere at
mid latitudes.
10-11
-------
Pacific Ocean (Cooper et al.. 2010; Tanimoto. 2009). The satellite trends over the
northern mid-latitudes are consistent with this picture as well (Ziemke et al., 2005).
Increases in tropospheric O3 concentrations in the Southern Hemisphere are also
likely due to increased anthropogenic NOX emissions, especially from biomass
burning (Fishman et al., 1991). Recent declines in summertime O3 concentrations
over Europe can be partly explained by decreases in O3 precursor emissions there
(Jonson et al.. 2005). while springtime increases at some European sites are likely
linked to changes in stratospheric dynamics (Ordonez et al.. 2007). Over Canada,
Fusco and Logan (2003) found that O3 depletion in the lowermost stratosphere may
have reduced the stratospheric flux of O3 into the troposphere by as much as 30%
from the early 1970s to the mid 1990s, consistent with the trends in ozonesondes
there.
Calculation of O3 Concentration Trends for the Recent Past
Simulations of trends in tropospheric O3 concentrations provide a means for testing
current knowledge of O3 processes and predicting with greater confidence trends in
future O3 concentrations. Time-dependent emission inventories of O3 precursors
have also been developed for 1850-2000 (Lamarque et al.. 2010) and for 1890-1990
(Van Aardenne et al.. 2001). These inventories allow for the calculation of changing
O3 concentration over time.
One recent multi-model study calculated an increase in the O3 concentration since
pre-industrial times of 8-14 DU, or about 30-70% (Gauss et al.. 2006). The large
spread in modeled estimates reveals the limitations in knowledge of processes in the
pristine atmosphere. Models typically overestimate the late nineteenth and early
twentieth century observations available in surface air and at mountain sites by 50-
100% (Lamarque et al.. 2005: Shindell et al.. 2003: Micklev et al.. 2001: Kiehl et al..
1999). Reconciling the differences between models and measurements will require
more accurate simulation of the natural sources of O3 (Micklev et al.. 2001) and/or
implementation of novel sinks such as bromine radicals, which may reduce
background O3 concentrations in the pristine atmosphere by as much as 30% (Yang
et al.. 2005c).
For the more recent past (since 1970), application of time-dependent emissions
reveals an equatorward shift in the distribution of tropospheric O3 in the Northern
Hemisphere due to the industrialization of societies at low-latitudes (Lamarque et al..
2005: Berntsen et al.. 2000). By constraining a model with historical (1950s-2000)
observations, Shindell and Faluvegi (2002) calculated a large increase of 8.2 DU in
tropospheric O3 abundance over polluted continental regions since 1950. This trend
is not captured in standard chemistry models, but is consistent with the change in
tropospheric O3 concentrations since pre-industrial times implied by the observations
from the late 1800s (Pavelin et al.. 1999: Marenco et al.. 1994).
10-12
-------
10.3.3.2 The Effect of Surface Albedo on Os Radiative Forcing
The Earth's surface albedo plays a role in O3 radiative forcing. Through most of the
troposphere, absorption of incoming shortwave solar radiation by O3 is small relative
to its absorption of outgoing longwave terrestrial radiation. However, over surfaces
characterized by high albedo (e.g., over snow, ice, or desert sand), incoming
radiation is more likely to be reflected than over darker surfaces, and the probability
that O3 will absorb shortwave solar radiation is therefore larger. In other words,
energy that would otherwise return to space may instead be retained in the
atmosphere. Several studies have shown that transport of O3 to the Arctic from mid-
latitudes leads to radiative forcing estimates greater than 1.0 W/m2 in the region,
especially in summer (Shindell et al., 2006; Liao et al., 2004b; Mickley et al, 1999).
Both the high surface albedo of the Arctic and the large solar zenith angles there
(which increase the path length of incoming sunlight) lead to strong shortwave
radiative forcing in the region. Because the Arctic is especially sensitive to radiative
forcing through the ice-albedo feedback, the large contribution in the shortwave solar
spectrum to the total radiative forcing in the region may be important.
10.3.3.3 The Effect of Vertical Distribution on Os Radiative Forcing
In the absence of feedbacks, O3 increments near the tropopause produce the largest
increases in surface temperature (Lacis et al.. 1990: Wang et al.. 1980). This is a
result of the colder temperature of the tropopause relative to the rest of the
troposphere and stratosphere. Since radiation emitted by the atmosphere is
approximately proportional to the fourth power of its temperature1, the colder the
added O3 is relative to the earth's surface, the weaker the radiation emitted and the
greater the "trapping" of longwave radiation in the troposphere.
10.3.3.4 Feedback Factors that Alter the Climate Response to
Changes in O3 Radiative Forcing
Estimates of radiative forcing provide a first-order assessment of the effect of
tropospheric O3 on climate. In the atmosphere, climate feedbacks and transport of
heat alter the sensitivity of Earth's surface temperature to addition of tropospheric
O3. Assessment of the full climate response to increases in tropospheric O3
concentrations requires use of a climate model to simulate these interactions.
Due to its short lifetime, O3 is heterogeneously distributed through the troposphere.
Sharp horizontal gradients exist in the radiative forcing of O3, with the greatest
radiative forcing since pre-industrial times occurring over the northern mid-latitudes
(more on this in Section 10.3.5 and Section 10.3.6). If climate feedbacks are
1 As described by the Stefan-Boltzmann law, an ideal blackbody-which the atmosphere approximates-absorbs at all wavelengths
and re-radiates proportional to the fourth power of its temperature.
10-13
-------
particularly powerful, they may obscure or even erase the correlation between
regional radiative forcing and climate response (Harvey, 2004; Boer and Yu, 2003).
The transport of heat through the atmosphere, though not technically a feedback, may
also weaken the correlation between radiative forcing and climate response. Several
model studies have reported that the horizontal pattern of surface temperature
response from 2000-2100 trends in predicted short-lived species (including O3)
closely matches the pattern from the trends in the long-lived greenhouse gases over
the same time period (Lew et al. 2008: Shindell et al.. 2008: Shindell et al. 2007).
This correspondence occurs even though the patterns of radiative forcing for the
short-lived and long-lived species differ substantially. In a separate paper, Shindell et
al. (2007) found that Arctic temperatures are especially sensitive to the mid-latitude
radiative forcing from tropospheric O3.
Other studies have found that the signature of warming due to tropospheric O3 does
show some consistency with the O3 radiative forcing. For example, Mickley et al.
(2004) examined the change in O3 concentrations since pre-industrial times and
found greater warming in the Northern Hemisphere than in the Southern Hemisphere
(+0.4°C versus +0.2°C), as well as higher surface temperatures downwind of Europe
and Asia and over the North American interior in summer. For an array of short-lived
species including O3, Shindell and Faluvegi (2009) found that radiative forcing
applied over northern mid-latitudes yield more localized responses due to local
cloud, water vapor, and albedo feedbacks than radiative forcing applied over the
tropics.
Climate feedbacks can also alter the sensitivity of surface temperature to the vertical
distribution of tropospheric O3. The previous section (Section 10.3.3.3) described the
greater effect of O3 added to the upper troposphere (near the tropopause) on radiative
forcing, relative to additions in the mid- to lower troposphere. However, warming
induced by increased O3 concentrations in the upper troposphere could stabilize the
atmosphere to some extent, limiting the transport of heat to the Earth's surface and
mitigating the effect of the added O3 on surface temperature (Joshi et al., 2003:
Christiansen, 1999). Hansen et al. (1997) determined that allowing cloud feedbacks
in a climate model meant that O3 enhancements in the mid-troposphere had the
greatest effect on surface temperature.
Finally, climate feedbacks can amplify or diminish the climate response of one
greenhouse gas relative to another. For example, Micklev et al. (2004) found a
greater temperature response to CO2 radiative forcing than to an O3 radiative forcing
of similar global mean magnitude, due in part to the relatively weak ice-albedo
feedback for O3 radiative forcing. Since CO2 absorbs in the same bands as water
vapor, CO2 radiative forcing saturates in the middle troposphere and is also shifted
toward the drier poles. A poleward shift in radiative forcing amplifies the ice-albedo
feedback in the case of CO2, and the greater mid-troposphere radiative forcing allows
for greater surface temperature response, relative to that for O3.
10-14
-------
10.3.3.5 Indirect Effects of Tropospheric O$ on the Carbon Cycle
A proposed indirect effect of tropospheric O3 on climate involves the carbon cycle.
By directly damaging plant life in ways discussed in Chapter 9, increases in
tropospheric O3 concentrations may depress the land-carbon sink of CO2, leading to
accumulation of CO2 in the atmosphere and ultimately warming of the Earth's
surface. Sitch et al. (2007) calculated that this indirect warming effect of O3 on
climate has about the same magnitude as the O3 direct effect. Their results suggest a
doubled sensitivity of surface temperatures to O3 radiative forcing, compared to
current model estimates.
A large array of additional indirect effects involving biospheric responses to
tropospheric O3 concentrations are possible. For example, increasing temperature
due to increases in tropospheric O3 concentration may alter biodiversity, species
migration, and consequent impacts on surface albedo. Such long-term feedbacks may
play an important role in the eventual climate response to changes in tropospheric O3
abundance, but a full evaluation of such long-term feedbacks on climate change is
outside the scope of this assessment.
10.3.4 Competing Effects of Os Precursors on Climate
Changes in concentrations of O3 precursors can affect the radiative balance of the
atmosphere through multiple (and sometimes competing) mechanisms. For example,
the O3 precursor CH4 is itself a powerful greenhouse gas. Ozone and its other
precursors also exert a strong control on the oxidizing capacity of the troposphere,
and so can affect the lifetime of gases such as CH4 (Derwent et al.. 2001). For
example, an increase in CO or VOCs would lead to a decrease in hydroxyl (OH)
concentrations. Since OH is a major sink for CH4, a decline in OH would lengthen
the CH4 lifetime, enhance the CH4 concentration, and amplify surface warming.
A rise in NOX emissions, on the other hand, could lead to an increase in OH in
certain locations, shortening the CH4 lifetime and causing surface cooling
(Fuglestvedt et al.. 1999). Ozone can itself generate OH through (1) photolysis
leading to excited oxygen atoms followed by reaction with water vapor and
(2) reaction with HO2.
Figure 10-4 shows the radiative forcing associated with a suite of anthropogenic
emissions, including O3 precursors (IPCC, 2007b). The emission-based radiative
forcing for CH4, which includes the CH4 effect on O3 production, is +0.9 W/m2, or
nearly double that of the CH4 abundance-based radiative forcing shown in
Figure 10-3. Figure 10-4 also shows a warming from anthropogenic CO and VOC
emissions of+0.27 W/m2 and a net cooling of-0.21 W/m2 for NOX emissions.
The net cooling for NOX occurs mainly due to the links between NOX and CH4.
Consistent with these results, Shindell and Faluvegi (2009) calculated positive
(+0.25 W/m2) radiative forcing from the increase in anthropogenic emissions of CO
and VOCs since pre-industrial times, as well as for CH4 (+1 W/m2). In contrast,
10-15
-------
Shindell and Faluvegi (2009) found negative (-0.29 W/m2) radiative forcing from
anthropogenic emissions of NOX. Other studies have found a near cancellation of the
positive O3 radiative forcing and the negative CH4 radiative forcing that arise from
an incremental increase in anthropogenic NOX emissions (Naik et al.. 2005; Fiore et
al.. 2002; Fuglestvedt et al., 1999). The net effect of aircraft NOX on climate is
especially complex (Isaksen et al.. 2001; Wild et al.. 2001). Stevenson (2004)
calculated that aircraft NOX leads to short-term net warming via O3 production in the
cool upper troposphere, but long-term net cooling because of CH4 loss.
OH production from O3 precursors can also affect regional sulfate air quality and
climate by increasing gas-phase oxidation rates of SO2. Using the A1B scenario in
the IPCC AR4, Unger (2006) reported that by 2030, enhanced OH from the A1B O3
precursors may increase surface sulfate aerosol concentrations by up to 20% over
India and China, relative to the present-day, with a corresponding increase in
radiative cooling over these regions. In this way, O3 precursors may impose an
indirect cooling via sulfate (Unger, 2006).
Taken together, these results point out the need for careful assessment of net
radiative forcing involving multiple pollutants in developing climate change policy
(Unger et al.. 2008). Many studies point to CH4 as a particularly attractive target for
emissions control since CH4 is itself an important precursor of O3 (West et al.. 2007;
Fiore et al.. 2002). Fiore et al. (2002) found that reducing anthropogenic CH4
emissions by 50% would lead to a global negative (-0.37 W/m2) radiative forcing,
mostly from CH4. In later research, Fiore et al. (2008) reported that CH4 reductions
would most strongly affect tropospheric O3 column amounts in regions of strong
down welling from the upper troposphere (e.g., around 30°N) and in regions of NOX-
saturated conditions.
The magnitude of the radiative forcing from the change in tropospheric O3
abundance since the pre-industrial era is uncertain. This uncertainty derives in part
from the scarcity of early measurements and in part from limited knowledge
regarding processes in the natural atmosphere. As noted previously, the IPCC AR4
reports a radiative forcing of 0.35 W/m2 from the change in tropospheric O3
abundance since 1750 (Forster et al.. 2007). ranking it third in importance behind the
greenhouse gases CO2 and CH4. The O3 radiative forcing could, in fact, be as large
as 0.7 W/m2, if reconstructions of pre-industrial and mid-20th century O3
concentrations based on the measurement record are valid (Shindell and Faluvegi.
2002; Mickley et al.. 2001). In any event, Unger et al. (2010) showed that present-
day O3 radiative forcing can be attributed to emissions from many economic sectors,
including on-road vehicles, household biofuel, power generation, and biomass
burning. As much as one-third of the radiative forcing from the 1890 to 1990 change
in tropospheric O3 concentration could be due to increased biomass burning (Ito et
al.. 2007a).
These calculated radiative forcing estimates can be compared to those obtained from
satellite data. Using data from TOMS, Worden et al. (2008) estimated a reduction in
clear-sky outgoing longwave radiation of 0.48 W/m2 by O3 in the upper troposphere
over oceans in 2006. This radiative forcing includes contributions from both
10-16
-------
anthropogenic and natural sources of O3. Assuming that the concentration of O3 has
roughly doubled since pre-industrial times (Gauss et al, 2006), the total O3 radiative
forcing estimated with TOMS is consistent with that obtained from models
estimating just the anthropogenic contribution.
Components of radiative forcing for principal emissions
Black carbon
SO2
Organic carbon
Mineral dust
Aerosols
Aircraft
Black carbon
(snow albedo)
Organic carbon
(direct)
Cloud albedo effe
Surface* albedo
(tantPuse)
Land use
Solar irradiance
-O.5
O 0.5
Radiative Forcing (W rrr2)
1.5
Note: Values represent radiative forcing in 2005 due to emissions and changes since 1750. (S) and (T) next to gas species
represent stratospheric and tropospheric changes, respectively.
Source: Reprinted with permission of Cambridge University Press (IPCC. 2007b).
Figure 10-4 Components of radiative forcing for emissions of principal gases,
aerosols, aerosol precursors, and other changes.
10-17
-------
10.3.5 Calculating Radiative Forcing and Climate Response to Past Trends
in Tropospheric O3 Concentrations
Calculation of the climate response to the O3 radiative forcing is challenging due to
complexity of feedbacks, as mentioned in Section 10.3.2.2 and Section 10.3.3.4.
In their modeling study, Micklev et al. (2004) reported a global mean increase of
0.28°C since pre-industrial times, with values as large as 0.8°C in continental
interiors. For the time period since 1870, Hansen et al. (2005) estimated a much
smaller increase in global mean surface temperature (0.11°C), but they implemented
1880s anthropogenic emissions in their base simulation and also took into account
trends in both stratospheric and tropospheric O3 concentrations. The modeled decline
of lower stratospheric O3 concentrations, especially over polar regions, cooled
surface temperatures in this study, counteracting the warming effect of increasing
tropospheric O3 concentrations.
Figure 10-5 shows the Hansen et al. (2005) results as reported in Shindell et al.
(2006). In that figure, summertime O3 has the largest radiative effect over the
continental interiors of the Northern Hemisphere. Shindell et al. (2006) estimated
that the change in tropospheric O3 concentration over the 20th century could have
contributed about 0.3°C to annual mean Arctic warming and as much as 0.4-0.5°C
during winter and spring. Over eastern China, Chang et al. (2009) calculated a
surface temperature increase of 0.4°C to the 1970-2000 change in tropospheric O3
concentration. It is not clear, however, to what degree regional changes in O3
concentration influenced this response, as opposed to more global changes.
10-18
-------
Annual surface air temperature^
Annual radiative forcing
-1.1 -.9 -.7 ^5 -.3 -.1 .1 .3 .5 .7 .9 1.1 0 .1 .2 .3 .4 .5 .6 .7
Summer (JJA) surface air temperature .10 Winter (DJF) surface air temperature
^9
••}
v Ar^-
|/' l> O
,>,
-1.1 -.9 -.7 -.5 -.3 -.1 .1 .3 .5 .7 .9 1.5 -1.1 -.9 -.7 ^5 -.3 -.1 .1 .3 .5 .7 .9 1.4
Note: This figure includes the input radiative forcing (W/m2), as computed by the NASA GISS chemistry-climate model. Values are
surface temperature trends for the annual average (top left), June-August (bottom left), and December-February (bottom right)
and annual average tropopause instantaneous radiative forcing from 1880 to 1990 (top right). Temperature trends greater than
about 0.1 °C are significant over the oceans, while values greater than 0.3°C are typically significant over land, except for northern
middle and high latitudes during winter where values in excess of about 0.5°C are significant. Values in the top right corner give
area-weighted global averages in the same units as the plots.
Source: Reprinted with permission of American Geophysical Union (Shindell et al.. 2006).
Figure 10-5 Ensemble average 1900-2000 radiative forcing and surface
temperature trends (°C per century) in response to tropospheric
O3 concentration changes.
10.3.6 Calculating Radiative Forcing and Climate Response to Future
Trends in Tropospheric O3 Concentrations
Future trends in tropospheric O3 concentrations depend in large part on what
pathways in energy technology the world's societies will follow in coming decades.
The trends in O3 concentration will also depend on the changes in a suite of climate-
sensitive factors, such as the water vapor content of the atmosphere. This section
describes the following issues: (1) projected trends in the anthropogenic emissions of
O3 precursors; (2) the effects of these emissions on the tropospheric O3
concentrations; (3) the effects of changing climate on tropospheric O3
concentrations; and (4) radiative forcing and climate response to 21st century trends
in tropospheric O3 concentrations.
10-19
-------
10.3.6.1 Emissions of Anthropogenic O$ Precursors Across the 21st
Century
The IPCC SRES effort devised scenarios for short-lived O3 precursors as well as the
well-mixed greenhouse gases including NOX, CO, and VOCs (IPCC, 2000). Using
the IMAGE socioeconomic model, Streets et al. (2004) provided speciation for NOX
and VOCs and allocated the trends in emissions over 17 regions and 8 economic
sectors for the 2000-2050 time period. The worst-case IPCC scenario, A2, features
continued dependence on fossil fuels, rapid population growth, and little sharing of
technology between developed and developing nations. By 2100 in this scenario,
global NOX, CO and CH4 emissions increase by a factor of 3.5, 2.6, and 2.9,
respectively, relative to 2000 (IPCC. 2000). Most of these increases in emissions
occur over developing countries. For example over Asia, NOX emissions in the A2
scenario increase by more than a factor of four by 2100. The more moderate A1B
scenario has global NOX and CO emissions increasing by 25% and 90%, respectively
by 2100, but global CH4 emissions decreasing by 10%. In the Bl scenario, with its
emphasis on clean and efficient technologies, global emissions of NOX, CO, and
CH4 all decrease by 2100, relative to the present day (-40%, -60%, and -30%,
respectively).
Other emissions scenarios have been recently developed to describe trends in the
short-term (up to 2030). The Current Legislation (CLE) scenario provides trends
consistent with existing air quality regulations; the Maximum Feasible Reduction
(MFR) scenario seeks to reduce emissions of O3 precursors to the maximum extent
possible. Emission source changes relative to the present day for CLE, MFR, and A2
are given in Stevenson et al. (2006).
For the Fifth Assessment Report (IPCC AR5), a new set of climate futures has been
developed: the Representative Concentration Pathways (RCPs) (Moss et al., 2010).
The RCPs will explore for the first time approaches to climate change mitigation.
The RCPs are designed to achieve radiative forcing targets of 2.6, 4.5, 6.0 and
8.5 W/m2 by 2100, and have been designated RCP 2.6, RCP 4.5, RCP 6.0, and RCP
8.5, respectively (RCP 2.6 is also known as RCP3-PD.) The trends in O3 precursors
for the RCP scenarios were determined by climate policies implicit in each scenario
and by plausible assumptions regarding future air quality regulations. These
scenarios were chosen to map the wide range of climate outcomes presented in the
literature and represent only four of many possible scenarios that would lead to the
specific radiative forcing targets; a wide range of socioeconomic conditions could be
consistent with each radiative forcing pathway (Moss et al., 2010). Therefore, they
should not be interpreted as forecasts of future conditions, but rather as plausible
climate and socio-economic futures.
Plots and comparisons of the RCP trends are available on the RCP website (RCP.
2009). In all RCPs, global anthropogenic NOX emissions decline 30-50% during the
21st century, though RCP 8.5 shows a peak during the 2020s at a value ~15% greater
than that of 2000. Global anthropogenic VOC and CO emissions are relatively flat
during the 2000-2050 time range, and then decline by 30-50% by the end of the
10-20
-------
century. For CH4, global mean emission trends for the four RCP projections differ
substantially across the 21st century, with RCP 8.5 showing a tripling of emissions
by 2100, and RCP 2.6 showing the emissions cut by half in this time range. RCP 4.5
and 6.0 show a peak in CH4 emissions in the middle of the century before dropping
by the end of the century to just below 2000 emission levels. All these global trends,
however, contain some regional variation. For example, Asian emissions of both
NOX and VOCs show large increases in the near term (2030s to 2050s).
10.3.6.2 Impact of 21st Century Trends in Emissions on
Tropospheric Os Concentrations
Due to its short lifetime, tropospheric O3 concentrations will respond readily to
changes in anthropogenic emissions of O3 precursors. As shown in Table 10-1. a
recent multi-model study found increases in the tropospheric O3 concentration of
15% and 6% for the IPCC A2 and CLE scenarios respectively for the 2000-2030
time period, and a decrease for the MFR scenario of 5% (Stevenson et al.. 2006).
These results indicate that the growth in tropospheric O3 concentrations between
2000 and 2030 could be reduced or even reversed, depending on emission controls.
For the relatively moderate A1B emissions scenario over the 2000-2050 time period,
Wu et al. (2008a) calculated a change in O3 concentration of about 20%.
As noted above, the RCP scenarios show large variations in their future projections
of global mean CH4 emissions, but mainly declines in the emissions of the other O3
precursors across the 21st century. In one of the first efforts to assess the effect of
these emission trends on global O3 abundances, Lamarque et al. (2011) found that
the large CH4 increase in the RCP 8.5 scenario would drive a 15% enhancement of
the tropospheric O3 abundance by 2100, relative to the present-day, leading to a
global mean radiative forcing of+0.2 W/m2. By contrast, the global O3 abundance
would decrease in the other three RCPs, with declines in radiative forcing ranging
from-0.07 to-0.2 W/m2.
10-21
-------
Table 10-1 Changes in anthropogenic emissions, CH4 and tropospheric O3
concentrations between 2000 and 2030, and the associated
tropospheric O3 radiative forcing for three scenarios.
Scenario
Percent change in NOX emissions
Percent change in CO emissions
Percent change in ChU concentration
Percent change in
tropospheric O3 concentration
Radiative forcing due to
Os concentration change13 (W/m2)
IPCCA2a
+96%
+62%
+23%
+ 15%
0.3
Current Legislation
(CLE)a
+ 18%
-16%
+ 19%
+6%
0.18
Maximum Feasible Reduction
(MFR)a
-53%
-53%
0%
-5%
-0.05
aValues are ensemble means.
""Includes radiative forcing due to corresponding CH4 change.
Source: Adapted from Stevenson et al. (2006).
10.3.6.3 Impact of 21st Century Climate on Tropospheric O$
Concentrations
For the time period from the 1800s to the present-day, most of the increase in the
concentration of tropospheric O3 can be traced to changing emissions. Model studies
show that climate change so far has likely had little effect on the tropospheric O3
concentrations (e.g.. Grenfell et al.. 2001). In the future, however, climate change is
expected to bring large changes in a suite of variables that could affect O3
production, loss, and transport. For example, increased water vapor in a warming
atmosphere is expected to enhance OH concentrations, which in remote, NOx-poor
regions will accelerate O3 loss rates (Johnson et al.. 1999).
In the 2050s A1B climate, Wu et al. (2008b) calculated a 5 ppb decrease in surface
O3 concentrations over oceans. A rise in temperatures will also likely promote
emissions of isoprene, an important biogenic precursor of O3. Model studies have
calculated 21st-century increases in isoprene emissions ranging from 25-50%,
depending on climate scenario and time horizon (Wu et al.. 2008a and references
therein). These studies however did not take into account the effects of changing
climate and CO2 concentration on vegetation extent, which could have large
consequences for biogenic emissions (Heald et al.. 2008: Sanderson et al.. 2003).
In any event, enhanced isoprene emissions will increase O3 concentrations in VOC-
limited regions, but decrease O3 concentrations in NOx-limited regions (Wu et al..
2008a: Pyle et al.. 2007: Sanderson et al.. 2003). Convection frequencies and
lightning flash rates will also likely change in a changing climate, with consequences
for lightning NOX emissions and O3 concentrations in the upper troposphere (Sinha
andToumi. 1997: Price and Rind. 1994). While Wu et al. (2008a) calculated an
10-22
-------
increase in lightning NOX by 2050 due to enhanced deep convection, Jacobson and
Streets (2009) projected a decrease in lightning NOX due to a declining cloud ice in
their future atmosphere. Finally, changes in transport processes will almost certainly
accompany global climate change. For the 2050 A1B climate, Wu et al. (2008b)
showed that flattening of the meridional temperature gradient in a warming world
would lead to slower intercontinental transport of tropospheric O3. For the
A2 climate in 2100, Zeng and Pyle (2003) projected an 80% increase in the flux of
stratospheric O3 into the troposphere, relative to the present-day.
Taken together, these climate-driven processes could have appreciable effects on the
concentration and distribution of tropospheric O3. As shown in Wu et al. (2008b),
model projections of the change in O3 concentration due solely to future climate
change range from -12% to +3%, depending on the model, scenario, and time
horizon.
10.3.6.4 Radiative Forcing and Climate Response from 21st Century
Trends in Tropospheric O3 Concentrations
In the near term (2000-2030), Stevenson et al. (2006) estimated an O3 radiative
forcing of near zero for MFR, 0.18 W/m2 for CLE, and 0.3 W/m2 for the A2 scenario
(Table 10-1). Menon et al. (2008), following the moderate A1B scenario, calculated a
radiative forcing of 0.12 W/m2 from the 2000-2030 change in tropospheric O3
concentrations, about the same as that derived by Stevenson et al. (2006) for the CLE
scenario. Over the longer term (2000 to 2100) for the A1B scenario, Gauss et al.
(2003) reported large positive radiative forcing (0.40 to 0.78 W/m2) due to the
change in tropospheric O3 concentrations, as shown in Figure 10-6. Normalized
radiative forcing for these model calculations fell within a relatively narrow range,
0.032 to 0.040 W/m2 DU, indicating that the largest uncertainty lies in the model-
calculated changes in O3 concentration. Applying the A2 scenario, Chen et al.
(2007b) estimated a global mean radiative forcing of 0.65 W/m2 from tropospheric
O3 by 2100, consistent with the Gauss et al. (2003) results. These studies took into
account only the effect of changing emissions on tropospheric O3 concentrations.
In their calculations of the 2000-2100 radiative forcing from O3 in the A2 scenario,
Liao et al. (2006) found that inclusion of climate effects on tropospheric O3 reduced
their radiative forcing estimate by 20%.
10-23
-------
0.90
0.80
0.70
'E 0.60
g1 0.50
u
5
, 0.40
0.30
0.20
0.10
0.00
SW L
SW t + s LW t+sB INet t+s
LLAQ UI01 UCI IASB KNMI UCAM VICZ1 MOZ2 HGIS UKMO UI02
Note: Shown are the components of radiative forcing in W/m2. SW = shortwave component; LW = longwave component; Net = total
radiative forcing; t = tropospheric O3 changes only; and t+s = both tropospheric and stratospheric changes.
Source: Reprinted from Gauss et al. (2003), American Geophysical Union.
Figure 10-6 Global mean radiative forcing estimates calculated by a set of
models for the 2000-2100 change in tropospheric O3
concentrations.
Several studies have included tropospheric O3 in their investigations of the response
in the future atmosphere to a suite of short-lived species (e.g., Lew et al.. 2008:
Shindell et al.. 2008: Shin dell et al.. 2007). Few studies, however, have calculated the
climate response to changes in tropospheric O3 concentrations alone in the future
atmosphere. For the A2 atmosphere, Chen et al. (2007b) estimated a global mean
surface temperature increase of+0.34°C by 2100 in response to the change in O3
concentration. The largest temperature increases in this study, as much as 5°C,
occurred over the populous regions of Asia and the Middle East and downwind of
biomass burning regions in South Africa and South America.
10-24
-------
10.4 UV-B Shielding Effects and Tropospheric O3
10.4.1 Background
UV radiation emitted from the Sun contains sufficient energy when it reaches the
Earth to break (photolyze) chemical bonds in molecules, thereby leading to damaging
effects on living organisms and materials. Atmospheric O3 plays a crucial role in
reducing exposure to solar UV radiation at the Earth's surface. Stratospheric O3 is
responsible for the majority of this shielding effect, as approximately 90% of total
atmospheric O3 is located there over mid-latitudes (Kar et al.. 2010: Crist et al..
1994). Investigation of the supplemental shielding of UV-B radiation provided by
tropospheric O3 is necessary for quantifying UV-B exposure and the incidence of
related human health effects, ecosystem effects, and materials damage. The role of
tropospheric O3 in shielding of UV-B radiation is discussed in this section.
10.4.2 Human Exposure and Susceptibility to Ultraviolet Radiation
The factors that potentially influence UV radiation exposure were discussed in detail
in Chapter 10 of the 2006 O3 AQCD (U.S. EPA. 2006b) and are summarized here.
These factors included outdoor activity, occupation, age, sex, geography, and
protective behavior. Outdoor activity and occupation both influenced the amount of
time people spend outdoors during daylight hours, the predominant factor for
exposure to solar UV radiation. Age and sex were found to be factors that influence
human exposure to UV radiation, particularly by influencing other factors of
exposure such as outdoor activity and risk behavior. Studies indicated that females
generally spent less time outdoors and, consequently, had lower UV radiation
exposure on average compared to males. Geography influences the degree of solar
UV flux to the surface, and hence exposure to UV radiation. Higher solar flux at
lower latitudes increased the annual UV radiation dose for people living in southern
states relative to northern states. Altitude was also found to influence personal
exposure to UV radiation. Protective behaviors such as using sunscreen, wearing
protective clothing, and spending time in shaded areas were shown to reduce
exposure to UV radiation. Given these and other factors that potentially influence
UV radiation exposure, the 2006 O3 AQCD (U.S. EPA. 2006b) listed the following
subpopulations potentially at risk for higher exposures to UV radiation:
• Individuals who engage in high-risk behavior (e.g., sunbathing);
• Individuals who participate in outdoor sports and activities;
• Individuals who work outdoors with inadequate shade (e.g., farmers,
construction workers, etc.);
• Individuals living in geographic areas with higher solar flux including lower
latitudes (e.g., Honolulu, HI) and higher altitudes (e.g., Denver, CO).
10-25
-------
The risks associated with all these factors are, of course, highly dependent on season
and region (Sliney and Wengraitis, 2006).
10.4.3 Human Health Effects due to UV-B Radiation
Chapter 10 of the 2006 O3 AQCD (U.S. EPA. 2006b) covered in detail the human
health effects associated with solar UV-B radiation exposure. These effects include
erythema, skin cancer, ocular damage, and immune system suppression. These
adverse effects, along with protective effects of UV radiation through increased
production of vitamin D are summarized in this section. For additional details, the
reader is referred to Chapter 10 of the 2006 O3 AQCD (U.S. EPA. 2006b) and
references therein.
The most conspicuous and well-recognized acute response to UV radiation is
erythema, or the reddening of the skin. Erythema is likely caused by direct damage to
DNA by UV radiation. Many studies discussed in Chapter 10 of the 2006 O3 AQCD
(U.S. EPA. 2006b) found skin type to be a significant risk factor for erythema. Skin
cancer is another prevalent health effect associated with UV radiation. Exposure to
UV radiation is considered to be a major risk factor for all forms of skin cancer.
Ocular damage from UV radiation exposure includes effects on the cornea, lens, iris,
and associated epithelial and conjunctival tissues. The region of the eye affected by
exposure to UV radiation depends on the wavelength of the incident UV radiation.
Depending on wavelength, common health effects associated with UV radiation
include photokeratitis (snow blindness; short wavelengths) and cataracts (opacity of
the lens; long wavelengths).
Experimental studies reviewed in Chapter 10 of the 2006 O3 AQCD (U.S. EPA,
2006b) have shown that exposure to UV radiation may suppress local and systemic
immune responses to a variety of antigens. Results from controlled human exposure
studies suggest that immune suppression induced by UV radiation may be a risk
factor contributing to skin cancer induction. There is also evidence that UV radiation
has indirect involvement in viral oncogenesis through the human papillomavirus,
dermatomyositis, human immunodeficiency virus, and other forms of
immunosuppression.
A potential health benefit of increased UV-B exposure relates to the production of
vitamin D in humans. Most humans depend on sun exposure to satisfy their
requirements for vitamin D. Vitamin D deficiency can cause metabolic bone disease
among children and adults, and also may increase the risk of many common chronic
diseases, including type I diabetes mellitus and rheumatoid arthritis. Substantial in
vitro and toxicological evidence also support a role for vitamin D activity against the
incidence or progression of various forms of cancer. In some studies, UV-B related
production of vitamin D had potential beneficial immunomodulatory effects on
multiple sclerosis, insulin-dependent diabetes mellitus, and rheumatoid arthritis.
More details on UV-B protective studies are provided in Chapter 10 of the 2006 O3
AQCD (U.S. EPA. 2006b).
10-26
-------
In establishing guidelines on limits of exposure to UV radiation, the International
Commission on Non-Ionizing Radiation Protection (ICNIRP) agreed that some low-
level exposure to UV radiation has health benefits (ICNIRP, 2004). However, the
adverse health effects of higher UV exposures necessitated the development of
exposure limits for UV radiation. The ICNIRP recognized the challenge in
establishing exposure limits that would achieve a realistic balance between beneficial
and adverse health effects. As concluded by ICNIRP (2004). "[t]he present
understanding of injury mechanisms and long-term effects of exposure to [UV
radiation] is incomplete, and awaits further research."
10.4.4 Ecosystem and Materials Damage Effects Due to UV-B Radiation
A 2009 progress report on the environmental effects of O3 depletion from the UNEP,
Environmental Effects Assessment Panel (UNEP, 2009) lists many ecosystem and
materials damage effects from UV-B radiation. An in-depth assessment of the global
ecosystem and materials damage effects from UV-B radiation per se is out of the
scope of this assessment. However, a brief summary of some mid-latitude effects is
provided in this section to provide context for UV-B related issues pertaining to
tropospheric O3. The reader is referred to the UNEP report (UNEP, 2009) and
references therein for further details. All of these UV-B related ecosystem and
materials effects can also be influenced by climate change through temperature and
other meteorological alterations, making quantifiable predictions of UV-B shielding
effects difficult.
Terrestrial ecosystem effects from increased UV-B radiation include reduced
plant productivity and plant cover, changes in biodiversity, susceptibility to infection,
and increases in natural UV protective responses. In general, however, these effects
are small for moderate UV-B increases at mid-latitudes. A field study on wheat in
southern Chile found no substantial changes in crop yield with moderate increases in
UV-B radiation (Calderini et al.. 2008). Similarly, field studies on silver birch
(Betula pendula) in Finland found no measurable effects in photosynthetic function
with increases in UV-B radiation (Aphalo et al.. 2009). Subtle, but important,
changes in habitat and biodiversity have also been linked to increases in UV-B
radiation (Mazza et al.. 2010: Obara et al.. 2008: Wahl. 2008). Some plants have
natural coping mechanisms for dealing with changes in UV-B radiation (Favory et
al.. 2009: Jenkins. 2009: Brown and Jenkins. 2008: loki et al.. 2008). but these
defenses may have costs in terms of reduced growth (Snell et al.. 2009: Clarke and
Robinson. 2008: Semerdjieva et al.. 2003: Phoenix et al.. 2000).
Aquatic ecosystem effects from increased UV-B radiation include sensitivity in
growth, immune response, and behavioral patterns of aquatic organisms. One study
looking at coccolithophores, an abundant phytoplankton group, found a 25%
reduction in cellular growth with UV-B exposure (Gao et al.. 2009a). Exposure to
relevant levels of UV-B radiation has been shown to modify immune response, blood
chemistry, and behavior in certain species offish (Markkula et al.. 2009: Holtbv and
10-27
-------
Bothwell, 2008; Jokinen et al., 2008). Adverse effects on growth and development
from UV-B radiation have also been observed for amphibians, sea urchins, mollusks,
corals, and zooplankton (Garcia et al., 2009b; Romansic et al., 2009; Croteau et al.,
2008b; Croteau et al.. 2008a; Marquis et al.. 2008; Marquis and Miaud. 2008; Oromi
et al., 2008). Increases in the flux of UV-B radiation may also result in an increase in
the catalysis of trace metals including mercury, particularly in clear oligotrophic
lakes with low levels of dissolved organic carbon to stop the penetration of UV-B
radiation (Schindler et al.. 1996). This could then alter the mobility of trace metals
including the potential for increased mercury volatilization and transport within and
among ecosystems.
Biogeochemical cycles, particularly the carbon cycle, can also be influenced by
increased UV-B radiation. A study on high latitude wetlands found UV-induced
increases in CO2 uptake through soil respiration (Haapala et al.. 2009) while studies
on arid terrestrial ecosystems found evidence for UV-induced release of CO2 through
photodegradation of above-ground plant litter (Brandt et al., 2009; Henry et al., 2008;
Caldwell et al., 2007; Zepp et al., 2007). Changes in solar UV radiation may also
have effects on carbon cycling and CO2 uptake in the oceans (Brewer and Peltzer,
2009; Meador et al.. 2009; Fritz et al.. 2008; Zepp et al.. 2008; Hader et al.. 2007) as
well as release of dissolved organic matter from sediment and algae (Mayer et al.,
2009; Riggsbee et al., 2008). Additional studies showing effects on these and
additional biogeochemical cycles including the water cycle and halocarbon cycle can
be found in the UNEP report (UNEP, 2009) and references therein.
Materials damage from increased UV-B radiation include UV-induced
photodegradation of wood (Kataoka et al.. 2007) and plastics (Pickett et al.. 2008).
These studies and others summarizing photo-resistant coatings and materials
designed to reduce photodegradation of materials are summarized in the UNEP
report (UNEP. 2009) and references therein.
The ecosystem, carbon cycle, and materials effects described in this section are for
UV-B exposure in general. Only a small fraction of these effects would be offset by
incremental decreases in UV-B exposure resulting from increases in tropospheric O3
concentrations. Attribution of UV-B shielding effects to changes in tropospheric O3
concentrations is a highly complex problem as discussed in the next section.
10.4.5 UV-B Shielding Effects Associated with Changes in Tropospheric
O3 Concentrations
There are multiple complexities in attempting to quantify the relationship between
changes in tropospheric O3 concentrations and UV-B exposure. The 2006 O3 AQCD
(U.S. EPA, 2006b) described a handful of studies addressing this relationship, but
none reported quantifiable effects of tropospheric O3 concentration fluctuations on
UV-B exposure at the surface. Further quantifying the relationship between UV-B
exposure and health or welfare effects is complicated by the uncertainties involved in
the selection of an action spectrum and appropriate characterization of dose
10-28
-------
(e.g., peak or cumulative levels of exposure, timing of exposures, etc.) The lack of
published studies that critically examined these issues together—that is the
incremental health or welfare effects attributable specifically to UV-B changes
resulting from changes in tropospheric O3 concentrations—lead to the prior
conclusion that the effect of changes in surface-level O3 concentrations on
UV-induced health outcomes could not be critically assessed within reasonable
uncertainty (U.S. EPA. 2006b).l
A recent study by Madronich et al. (2011) used CMAQ to estimate UV radiation
response to changes in tropospheric O3 concentrations under different control
scenarios projected out to 2020. This study focused on southeastern U.S. and
accounted for spatial and temporal variation in tropospheric O3 concentration
reductions, an important consideration since most controls are focused on reducing
O3 concentrations in populated urban areas. The contrasting control strategies
considered in this study included a historical scenario designed to meet an 84 ppb 8-h
daily max standard and a reduced scenario designed to bring areas predicted to
exceed a similarly designed 70 ppb standard into attainment. A biologically effective
irradiance was estimated by multiplying the modeled UV irradiance by a sensitivity
function (action spectrum) for the induction of nonmelanoma skin cancer in mice
corrected for human skin transmission, then integrating over UV wavelengths.
The average relative change in skin cancer-weighted surface UV radiation between
the two scenarios was 0.11 ±0.03% over June, July and August. Weighting by
population, this estimate increased to 0.19 ± 0.06%. Madronich et al. (2011) report
that their estimated UV radiation increment is an order of magnitude less than that
reported in an earlier study by Lutter and Wolz (1997) with the main reason for the
discrepancy coming from the overly-simplified uniform 10 ppb reduction in O3
concentrations assumed in the former study. Madronich et al. (2011) did not attempt
to link their predicted increase in UV radiation to a predicted increase in skin cancer
incidence, however, due to several remaining and substantial uncertainties.
Quantitatively estimating human health and welfare effects directly attributed to
changes in UV-B penetration resulting from changes in ground-level O3
concentrations will require both (a) a solid understanding of the multiple factors that
define the extent of exposure to UV-B, and (b) well-defined and quantifiable links
between UV-B exposure and human disease and welfare effects. Detailed
information does not exist regarding the relevant type (e.g., peak or cumulative) and
time period (e.g., developmental, lifetime, or current) of exposure, wavelength
dependency of biological responses, and inter-individual variability in UV resistance.
1 The reader is referred to the U.S. EPA 2003 Final Response to Court Remand (U.S. EPA. 2003) for detailed discussions of the
data and scientific issues associated with the determination of public health benefits resulting from the attenuation of UV-B by
surface-level O3.
10-29
-------
Although the UV-B related health effects attributed to marginal reductions in
tropospheric or ground-level O3 concentrations have not been directly assessed to
date, they would be expected to be small based on current information indicating a
negligibly small effect of potential future changes in tropospheric O3 concentrations
on ground-level UV-B radiation. In conclusion, the effect of changes in surface-level
O3 concentrations on UV-induced health and welfare outcomes cannot yet be
critically assessed within reasonable uncertainty.
10.5 Summary and Causal Determinations
10.5.1 Summary of the Effects of Tropospheric Os on Climate
Radiative forcing by a greenhouse gas or aerosol is a metric used to quantify the
change in balance between radiation coming into and going out of the atmosphere
caused by the presence of that substance. Tropospheric O3 is a major greenhouse gas
and radiative forcing agent; evidence from satellite data shows a sharp dip in the
outgoing infrared radiation in the 9.6 |j,m O3 absorption band. Models calculate that
the global average concentration of tropospheric O3 has doubled since the
pre-industrial era, while observations indicate that in some regions O3 may have
increased by factors as great as 4 or 5. These increases are tied to the rise in
emissions of O3 precursors from human activity, mainly fossil fuel consumption and
agricultural processes. Overall, the evidence supports a causal relationship
between changes in tropospheric O3 concentrations and radiative forcing.
While the developed world has successfully reduced emissions of O3 precursors in
recent decades, many developing countries have experienced large increases in
precursor emissions and these trends are expected to continue, at least in the near
term. Projections of radiative forcing due to changing O3 concentrations over the
21st century show wide variation, due in large part to the uncertainty of future
emissions of source gases. In the near-term (2000-2030), projections of O3 radiative
forcing range from near zero to +0.3 W/m2, depending on the emissions scenario
(Stevenson et al., 2006).
The impact of the tropospheric O3 change since pre-industrial times on climate has
been estimated to be about 25-40% of the anthropogenic CO2 impact and about 75%
of the anthropogenic CH4 impact according to the IPCC, ranking it third in
importance after CO2 and CH4. There are large uncertainties in the magnitude of the
radiative forcing estimate attributed to tropospheric O3, making the impact of
tropospheric O3 on climate more uncertain than the effect of the longer-lived
greenhouse gases. Furthermore, radiative forcing does not take into account the
climate feedbacks that could amplify or dampen the actual climate response (e.g.,
surface temperature change) that would result from a change in tropospheric O3
concentrations. Quantifying the change in surface temperature requires a complex
climate simulation in which all important feedbacks and interactions are accounted
10-30
-------
for. As these processes are not well understood or easily modeled, the surface
temperature response to a given radiative forcing is highly uncertain and can vary
greatly among models and from region to region within the same model. Even with
these these uncertainties, global climate models indicate that tropospheric O3 has
contributed to observed changes in global mean and regional surface temperatures.
As a result of such evidence presented in climate modeling studies, there is likely to
be a causal relationship between changes in tropospheric O3 concentrations
and effects on climate as quantified through surface temperature response.
Reduction of tropospheric O3 concentrations could therefore provide an important
means to slow climate change in addition to the added benefit of improving surface
air quality. However the precursors of O3 also have competing effects on the
greenhouse gas CH4, complicating emissions reduction strategies. A decrease in CO
or VOC emissions would enhance OH concentrations, shortening the lifetime of
CH4, while a decrease in NOX emissions could depress OH concentrations in certain
regions and lengthen the CH4 lifetime. Abatement of CH4 emissions would likely
provide the most straightforward means to address climate change since CH4 is itself
an important O3 precursor (West et al.. 2007: West et al.. 2006: Fiore et al.. 2002).
A reduction of CH4 emissions would also improve air quality on its own right. A set
of global abatement measures identified by West and Fiore (2005) could reduce CH4
emissions by 10% at a cost savings, decrease background O3 concentrations by about
1 ppb in the Northern Hemisphere summer, and lead to a global net cooling of
0.12 W/m2. West et al. (2007) explored further the benefits of CH4 abatement,
finding that a 20% reduction in global CH4 emissions would lead to greater cooling
per unit reduction in surface O3 concentration, compared to 20% reductions in VOCs
or CO.
Important uncertainties remain regarding the effect of tropospheric O3 on future
climate change. To address these uncertainties, further research is needed to:
(1) improve knowledge of the natural atmosphere; (2) interpret observed trends in O3
concentrations in the free troposphere and remote regions; (3) improve understanding
of the CH4 budget, especially emissions from wetlands and agricultural sources,
(4) understand the relationship between regional O3 radiative forcing and regional
climate change; and (5) determine the optimal mix of emissions reductions that
would act to limit future climate change.
10.5.2 Summary of UV-B Related Effects on Human Health, Ecosystems,
and Materials Relating to Changes in Tropospheric O3
Concentrations
UV radiation emitted from the Sun contains sufficient energy when it reaches the
Earth to break (photolyze) chemical bonds in molecules, thereby leading to damaging
effects on living organisms and materials. Atmospheric O3 plays a crucial role in
reducing exposure to solar UV radiation at the Earth's surface. Ozone in the
stratosphere is responsible for the majority of this shielding effect, as approximately
90% of total atmospheric O3 is located there over mid-latitudes. Ozone in the
10-31
-------
troposphere provides supplemental shielding of radiation in the wavelength band
from 280-315 nm, referred to as UV-B radiation. UV-B radiation has important
effects on human health and ecosystems, and is associated with materials damage.
EPA has found no published studies that adequately examine the incremental health
or welfare effects (adverse or beneficial) attributable specifically to changes in UV-B
exposure resulting from perturbations in tropospheric O3 concentrations. While the
effects are expected to be small, they cannot yet be critically assessed within
reasonable uncertainty. Overall, the evidence is inadequate to determine if a
causal relationship exists between changes in tropospheric O3 concentrations
and effects on health and welfare related to UV-B shielding.
10.5.3 Summary of O3 Causal Determinations
The evidence reviewed in this chapter describes the recent findings regarding the
climate and UV-B shielding effects of changes in tropospheric O3 concentrations.
Table 10-2 provides an overview of the causal determinations for each of the
categories evaluated including the effect of tropospheric O3 on radiative forcing,
climate change, and health and welfare effects related to UV-B shielding.
Table 10-2 Summary of O3 causal determinations for climate and
UV-B shielding effects.
Effects
Causal Determination
Radiative Forcing
Causal relationship
Climate Change
Likely to be a causal relationship
Health and Welfare Effects Related to UV-B Shielding Inadequate to determine if a causal relationship exists
10-32
-------
References
Aphalo. PJ: Vapaavuori. EM; de la Rosa. TM; Lehto. T. (2009). Does supplemental UV-B radiation affect gas
exchange and RuBisCO activity of Betula pendula Roth, seedlings grown in forest soil under greenhouse
conditions? Plant Ecol Divers 2: 37-43. http://dx.doi.org/10.1080/17550870902780299
Arrhenius. S. (1896). On the influence of carbonic acid in the air upon the temperature of the ground. Philos
Mag 41: 237-276.
Balis. PS: Zerefos. CS: Kourtidis. K; Bais. AF; Hofzumahaus. A; Kraus. A; Schmitt. R; Blumthaler. M; Gobbi.
GP. (2002). Measurements and modeling of photolysis rates during the photochemical activity and
ultraviolet radiation (PAUR) II campaign. J Geophys Res 107: 8138.
http://dx.doi.org/10.1029/2000JD000136
Berntsen. TK; Myhre. G: Stordal. F; Isaksen. ISA. (2000). Time evolution of tropospheric ozone and its
radiative forcing. J Geophys Res 105: 8915-8930. http://dx.doi.org/10.1029/1999JD901139
Boer. GJ: Yu. B. (2003). Climate sensitivity and response. Clim Dynam 20: 415-429.
http://dx.doi.org/10.1007/s00382-002-0283-3
Bony. S: Colman. R; Kattsov. VM; Allan. RP; Bretherton. CS: Dufresne. JL; Hall. A; Hallegatte. S: Holland.
MM; Ingram. W: Randall. DA; Soden. BJ; Tselioudis. G: Webb. MJ. (2006). How well do we understand
and evaluate climate change feedback processes? J Clim 19: 3445-3482.
Brandt. LA; Bohnet C: King. JY. (2009). Photochemically induced carbon dioxide production as a mechanism
for carbon loss from plant litter in arid ecosystems. J Geophys Res 114: G02004.
http://dx.doi.org/10.1029/2008jg000772
Brewer. PG: Peltzer. ET. (2009). Limits to marine life. Science 324: 347-348.
http://dx.doi.org/10.1126/science.1170756
Brown. BA; Jenkins. GI. (2008). UV-B signaling pathways with different fluence-rate response profiles are
distinguished in mature Arabidopsis leaf tissue by requirement for UVR8, HY5, and HYH. Plant Physiol
146: 576-588. http://dx.doi.org/10.1104/pp.107.108456
Bruhl. C: Crutzen. PJ. (1989). On the disproportionate role of tropospheric ozone as a filter against solar UV-B
radiation. Geophys Res Lett 16: 703-706. http://dx.doi.org/10.1029/GL016i007p00703
Calderini. DF; Lizana. XC: Hess. S: Jobet. CR; Zuniga. JA. (2008). Grain yield and quality of wheat under
increased ultraviolet radiation (UV-B) at later stages of the crop cycle. J Agr Sci 146: 57-64.
http://dx.doi.org/10.1017/S0021859607007447
Caldwell. MM; Bornman. JF; Ballare. CL; Flint. SD; Kulandaivelu. G. (2007). Terrestrial ecosystems,
increased solar ultraviolet radiation, and interactions with bother climate change factors. Photochem
Photobiol Sci 6: 252-266. http://dx.doi.org/10.1039/B700019g
Chang. W: Liao. H; Wang. H. (2009). Climate responses to direct radiative forcing of anthropogenic aerosols,
tropospheric ozone, and long-lived greenhouse gases in eastern China over 1951-2000. Adv Atmos Sci 26:
748-762. http://dx.doi.org/10.1007/s00376-009-9032-4
Chen. WT; Liao. H; Seinfeld. JH. (2007b). Future climate impacts of direct radiative forcing of anthropogenic
aerosols, tropospheric ozone, and long-lived greenhouse gases. J Geophys Res 112: D14209.
http://dx.doi.org/10.1029/2006JD008051
Christiansen. B. (1999). Radiative forcing and climate sensitivity: The ozone experience. Q J Roy Meteorol
Soc 125: 3011-3035. http://dx.doi.org/10.1002/qi.49712556011
10-33
-------
Clarke. LJ; Robinson. SA. (2008). Cell wall-bound ultraviolet-screening compounds explain the high
ultraviolet tolerance of the Antarctic moss, Ceratodon purpureus. New Phytol 179: 776-783.
http://dx.doi.0rg/10.llll/i.1469-8137.2008.02499.x
Cooper. OR; Parrish. DP; Stohl. A; Trainer. M; Nedelec. P; Thouret. V: Gammas. JP; Oltmans. SJ: Johnson.
BJ; Tarasick. D; Leblanc. T; Mcdermid. IS: Jaffe. D; Gao. R; Stith. J: Ryerson. T; Aikin. K; Campos. T;
Weinheimer. A; Avery. MA. (2010). Increasing springtime ozone mixing ratios in the free troposphere over
western North America. Nature 463: 344-348. http://dx.doi.org/10.1038/nature08708
Crist. KG: Carmichael. GR; John. K. (1994). UV-B exposure and atmospheric ozone - Evaluation of radiative
flux to changes in ambient ozone levels. J Hazard Mater 37: 527-538. http://dx.doi.org/10.1016/0304-
3894(93)E0096-K
Croteau. MC: Davidson. MA; Lean. DRS: Trudeau. VL. (2008a). Global increases in ultraviolet B radiation:
Potential impacts on amphibian development and metamorphosis [Review]. Physiol Biochem Zool 81: 743-
761. http://dx.doi.org/10.1086/591949
Croteau. MC: Martyniuk. CJ: Trudeau. VL; Lean. DRS. (2008b). Chronic exposure of rana pipiens tadpoles to
uvb radiation and the estrogenic chemical 4-tert-octylphenol. J Toxicol Environ Health A 71: 134-144.
Derwent. RG: Collins. WJ: Johnson. CE; Stevenson. DS. (2001). Transient behaviour of tropospheric ozone
precursors in a global 3-D CTM and their indirect greenhouse effects. Clim Change 49: 463-487.
Favory. JJ; Stec. A; Gruber. H; Rizzini. L; Oravecz. A; Funk. M; Albert. A; Cloix. C: Jenkins. GI; Oakelev. EJ;
Seidlitz. HK; Nagv. F; Ulm. R. (2009). Interaction of COP1 and UVR8 regulates UV-B-induced
photomorphogenesis and stress acclimation in Arabidopsis. EMBO J 28: 591-601.
http://dx.doi.0rg/10.1038/emboj.2009.4
Fiore. AM; Jacob. DJ: Field. BD; Streets. DG: Fernandes. SD; Jang. C. (2002). Linking ozone pollution and
climate change: The case for controlling methane. Geophys Res Lett 29: 1919.
http://dx.doi.org/10.1029/2002GL015601
Fiore. AM; West. JJ: Horowitz. LW: Naik. V; Schwartzkopf. MD. (2008). Characterizing the tropospheric
ozone response to methane emission controls and the benefits to climate and air quality. J Geophys Res
113: D08307. http://dx.doi.org/10.1029/2007JD009162
Fishman. J: Fakhruzzaman. K; Cros. B; Nganga. D. (1991). Identification of widespread pollution in the
Southern Hemisphere deduced from satellite analyses. Science 252: 1693-1696.
http://dx.doi.org/10.1126/science.252.5013.1693
Forster. P; Ramaswamv. V; Artaxo. P; Berntsen. T; Betts. R: Fahev. DW: Havwood. J: Lean. J: Lowe. DC:
Myhre. G: Nganga. J: Prinn. R; Raga. G: Schultz. M; Van Dorland. R. (2007). Changes in atmospheric
constituents and in radiative forcing. In Climate Change 2007: The physical science basis: Contribution of
Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (pp.
129-234). Cambridge, UK: Cambridge University Press, http://www.ipcc.ch/pdf/assessment-
report/ar4/wg l/ar4-wg 1 -chapter2.pdf
Fritz. JJ: Neale. PJ: Davis. RF; Peloquin. JA. (2008). Response of Antarctic phytoplankton to solar UVR
exposure: Inhibition and recovery of photosynthesis in coastal and pelagic assemblages. Mar Ecol Prog Ser
365: 1-16. http://dx.doi.org/10.3354/Meps07610
Fuglestvedt. JS: Berntsen. TK; Isaksen. ISA; Mao. H; Liang. XZ; Wang. WC. (1999). Climatic forcing of
nitrogen oxides through changes in tropospheric ozone and methane: Global 3D model studies. Atmos
Environ 33: 961-978. http://dx.doi.org/10.1016/81352-2310(98)00217-9
Fusco. AC: Logan. JA. (2003). Analysis of 1970-1995 trends in tropospheric ozone at Northern Hemisphere
midlatitudes with the GEOS-CHEM model. J Geophys Res 108: 4449.
http://dx.doi.org/10.1029/2002JD002742
Gao. KS: Ruan. ZX; Villafane. VE; Gattuso. JP: Helbling. EW. (2009a). Ocean acidification exacerbates the
effect of UV radiation on the calcifying phytoplankter Emiliania huxleyi. Limnol Oceanogr 54: 1855-1862.
10-34
-------
Garcia. TS: Paoletti. DJ: Blaustein. AR. (2009b). Correlated trait responses to multiple selection pressures in
larval amphibians reveal conflict avoidance strategies. Freshw Biol 54: 1066-1077.
http://dx.doi.0rg/10.llll/i.1365-2427.2008.02154.x
Gauss. M; Myhre. G: Isaksen. ISA; Grewe. V: Pitari. G: Wild. O: Collins. WJ: Dentener. FJ; Ellingsen. K;
Gohar. LK; Hauglustaine. DA; lachetti. D; Lamarque. JF; Mancini. E; Micklev. LJ; Prather. MJ; Pyle. JA;
Sanderson. MG: Shine. KP; Stevenson. DS: Sudo. K; Szopa. S: Zeng. G. (2006). Radiative forcing since
preindustrial times due to ozone change in the troposphere and the lower stratosphere. Atmos Chem Phys 6:
575-599.
Gauss. M; Myhre. G: Pitari. G: Prather. MJ: Isaksen. ISA; Berntsen. TK; Brasseur. GP; Dentener. FJ: Dement.
RG: Hauglustaine. DA; Horowitz. LW: Jacob. DJ: Johnson. M; Law. KS: Micklev. LJ: Miiller. JF;
Plantevin. PH; Pyle. JA; Rogers. HL; Stevenson. DS: Sundet. JK; Van Weele. M; Wild. O. (2003).
Radiative forcing in the 21st century due to ozone changes in the troposphere and the lower stratosphere. J
Geophys Res 108: 4292. http://dx.doi.org/10.1029/2002JD002624
Grenfell. JL; Shindell. DT; Koch. D; Rind. D. (2001). Chemistry-climate interactions in the Goddard Institute
for Space Studies general circulation model 2. New insights into modeling the preindustrial atmosphere. J
Geophys Res 106: 33435-33451.
Haapala. JK; Morsky. SK; Saarnio. S: Rinnan. R; Suokanerva. H; Kyr. E; Latola. K; Martikanen. PJ;
Holopainen. T; Silvola. J. (2009). Carbon dioxide balance of a fen ecosystem in northern Finland under
elevated UV-B radiation. Global Change Biol 15: 943-954. http://dx.doi.org/10.im/U365-
2486.2008.01785.x
Hader. DP; Kumar. HP; Smith. RC: Worrest. RC. (2007). Effects of solar UV radiation on aquatic ecosystems
and interactions with climate change. Photochem Photobiol Sci 6: 267-285.
http://dx.doi.org/10.1039/B700020k
Hansen. J: Sato. M; Ruedv. R; Nazarenko. L; Lacis. A; Schmidt. GA; Russell. G: Aleinov. I; Bauer. M; Bauer.
S: Bell. N; Cairns. B; Canuto. V: Chandler. M; Cheng. Y; Del Genio. A; Faluvegi. G: Fleming. E; Friend.
A; Hall. T; Jackman. C: Kellev. M; Kiang. N; Koch. D; Lean. J: Lerner. J: Lo. K; Menon. S: Miller. R;
Minnis. P; Novakov. T; Oinas. V: Perlwitz. J: Perlwitz. J: Rind. D: Romanou. A; Shindell. D: Stone. P;
Sun. S: Tausnev. N; Thresher. D; Wielicki. B; Wong. T; Yao. M; Zhang. S. (2005). Efficacy of climate
forcings. J Geophys Res 110: D18104. http://dx.doi.org/10.1029/2005JD005776
Hansen. JE; Sato. M; Ruedv. R. (1997). Radiative forcing and climate response. J Geophys Res 102: 6831-
6864. http://dx.doi.org/10.1029/96JD03436
Harvey. LDP. (2004). Characterizing the annual-mean climatic effect of anthropogenic CO2 and aerosol
emissions in eight coupled atmosphere-ocean GCMs. Clim Dynam 23: 569-599.
http://dx.doi.org/10.1007/s00382-004-0455-4
Heald. CL; Henze. DK; Horowitz. LW: Feddema. J: Lamarque. JF: Guenther. A; Hess. PG: Vitt. F; Seinfeld.
JH; Goldstein. AH; Fung. I. (2008). Predicted change in global secondary organic aerosol concentrations in
response to future climate, emissions, and land use change. J Geophys Res 113: D05211.
http://dx.doi.org/10.1029/2007jd009092
Held. IM; Soden. BJ. (2000). Water vapor feedback and global warming. Annual Review of Energy and the
Environment 25: 441-475.
Henry. HAL; Brizgys. K; Field. CB. (2008). Litter decomposition in a California annual grassland: Interactions
between photodegradation and litter layer thickness. Ecosystems 11: 545-554.
http://dx.doi.org/10.1007/sl0021-008-9141-4
Holland. MM; Bitz. CM. (2003). Polar amplification of climate change in coupled models. Clim Dynam 21:
221-232. http://dx.doi.org/10.1007/s00382-003-0332-6
Holtbv. LB; Bothwell. ML. (2008). Effects of solar ultraviolet radiation on the behaviour of juvenile coho
salmon (Oncorhynchus kisutch): Avoidance, feeding, and agonistic interactions. Can J Fish Aquat Sci 65:
701-711. http://dx.doi.org/10.1139/F08-013
10-35
-------
ICNIRP (International Commission on Non-Ionizing Radiation Protection). (2004). Guidelines on limits of
exposure to ultraviolet radiation of wavelengths between 180 nm and 400 nm (incoherent optical radiation).
In ICNIRP Guidelines. Oberschleissheim, Germany.
loki. M; Takahashi. S: Nakaiima. N; Fujikura. K; Tamaoki. M; Saii. H; Kubo. A; Aono. M; Kanna. M; Ogawa.
D; Fukazawa. J: Oda. Y; Yoshida. S: Watanabe. M; Hasezawa. S: Kondo. N. (2008). An unidentified
ultraviolet-B-specific photoreceptor mediates transcriptional activation of the cyclobutane pyrimidine
dimer photolyase gene in plants. Planta 229: 25-36. http://dx.doi.org/10.1007/s00425-008-0803-4
IPCC (Intergovernmental Panel on Climate Change). (2000). Special report on emissions scenarios: A special
report of Working Group III of the Intergovernmental Panel on Climate Change. Cambridge, UK:
Intergovernmental Panel on Climate Change; Cambridge University Press.
http ://www. grida.no/climate/ipcc/emission/
IPCC (Intergovernmental Panel on Climate Change). (2007b). Contribution of working group I to the fourth
assessment report of the intergovernmental panel on climate change. Cambridge, United Kingdom:
Cambridge University Press, http://www.ipcc.ch/publications and data/ar4/wg 1 /en/contents.html
IPCC (Intergovernmental Panel on Climate Change). (2007c). Summary for policymakers. In: The Physical
Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental
Panel on Climate Change. Cambridge, United Kingdom and New York, NY, USA: Cambridge University
Press, http://www.ipcc.ch/publications and data/ar4/wg2/en/spm.html
Isaksen. ISA; Berntsen. TK; Wang. WC. (2001). NOx emissions from aircraft: Its impact on the global
distribution of CH4 and O3 and on radiative forcing. Terr Atmos Ocean Sci 12: 63-78.
Ito. A; Sudo. K; Akimoto. H; Sillman. S: Penner. JE. (2007a). Global modeling analysis of tropospheric ozone
and its radiative forcing from biomass burning emissions in the twentieth century. J Geophys Res 112:
D24307. http://dx.doi.org/10.1029/2007JD008745
Jacobson. MZ; Streets. DG. (2009). Influence of future anthropogenic emissions on climate, natural emissions,
and air quality. J Geophys Res 114: DOS 118. http://dx.doi.org/10.1029/2008JD011476
Jaffe. D; Price. H; Parrish. D; Goldstein. A; Harris. J. (2003). Increasing background ozone during spring on
the west coast of North America. Geophys Res Lett 30: 1613. http://dx.doi.org/10.1029/2003GL017024
Jenkins. GI. (2009). Signal transduction in responses to UV-B radiation. Annu Rev Plant Biol 60: 407-431.
http://dx.doi.org/10.1146/annurev.arplant.59.032607.092953
Johnson. CE; Collins. WJ: Stevenson. DS: Derwent. RG. (1999). The relative roles of climate and emissions
changes on future tropospheric oxidant concentrations. J Geophys Res 104: 18631-18645.
http://dx.doi.org/10.1029/1999 JD900204
Jokinen. IE; Markkula. ES: Salo. HM; Kuhn. P; Nikoskelainen. S: Arts. MT; Browman. HI. (2008). Exposure
to increased ambient ultraviolet B radiation has negative effects on growth, condition and immune function
of juvenile Atlantic salmon (Salmo salar). Photochem Photobiol 84: 1265-1271.
http://dx.doi.0rg/10.llll/i.1751-1097.2008.00358.x
Jonson. JE; Simpson. D; Fagerli. H; Solberg. S. (2005). Can we explain the trends in European ozone levels?
Atmos Chem Phys 6: 51-66. http://dx.doi.org/10.5194/acp-6-51-2006
Joshi. M; Shine. KP; Ponater. M; Stuber. N; Sausen. R; Li. L. (2003). A comparison of climate response to
different radiative forcings in three general circulation models: Towards an improved metric of climate
change. Clim Dynam 20: 843-854. http://dx.doi.org/10.1007/s00382-003-0305-9
Kar. J; Fishman. J: Creilson. JK; Richter. A; Ziemke. J: Chandra. S. (2010). Are there urban signatures in the
tropospheric ozone column products derived from satellite measurements? Atmos Chem Phys 10: 5213-
5222. http://dx.doi.org/10.5194/acp-10-5213-2010
Kataoka. Y; Kiguchi. M; Williams. RS: Evans. PP. (2007). Violet light causes photodegradation of wood
beyond the zone affected by ultraviolet radiation. Holzforschung und Holzverwertung 61: 23-27.
http://dx.doi.org/10.1515/HF.2007.005
10-36
-------
Kiehl. JT; Schneider. TL; Portmann. RW: Solomon. S. (1999). Climate forcing due to tropospheric and
stratospheric ozone. J Geophys Res 104: 31239-31254. http://dx.doi.org/10.1029/1999JD900991
Lacis. AA; Wuebbles. DJ: Logan. JA. (1990). Radiative forcing of climate by changes in the vertical
distribution of ozone. J Geophys Res 95: 9971-9981. http://dx.doi.org/10.1029/JD095iD07p09971
Lamarque. JF; Bond. TC: Eyring. V; Granier. C: Heil. A; Klimont. Z; Lee. D; Liousse. C: Mieville. A; Owen.
B; Schultz. MG: Shindell. D; Smith. SJ: Stehfest. E; Van Aardenne. J: Cooper. OR; Kainuma. M;
Mahowald. N; Mcconnell. J. R.: Naik. V; Riahi. K; Van Vuuren. DP. (2010). Historical (1850-2000)
gridded anthropogenic and biomass burning emissions of reactive gases and aerosols: Methodology and
application. Atmos Chem Phys Discuss 10: 4963-5019. http://dx.doi.org/10.5194/acpd-10-4963-2010
Lamarque. JF; Hess. P; Emmons. L; Buja. L; Washington. W: Granier. C. (2005). Tropospheric ozone
evolution between 1890 and 1990. J Geophys Res 110: D08304. http://dx.doi.org/10.1029/2004JD005537
Lamarque. JF: Kyle. GP; Meinshausen. M; Riahi. K; Smith. SJ: Vuuren. DP; Conlev. AJ; Vitt. F. (2011).
Global and regional evolution of short-lived radiatively-active gases and aerosols in the Representative
Concentration Pathways. Clim Change 109: 191-212. http://dx.doi.org/10.1007/sl0584-011-0155-0
Lelieveld. J: van Aardenne. J: Fischer. H; de Reus. M; Williams. J: Winkler. P. (2004). Increasing ozone over
the Atlantic Ocean. Science 304: 1483-1487. http://dx.doi.org/10.1126/science.1096777
Lenoble. J. (1993). Atmospheric radiative transfer. Hampton, VA: A. Deepak Publishing.
http://www.worldcat.org/title/atmospheric-radiative-transfer/oclc/27769441
Lew. H; Schwaarzkopf. MD; Horowitz. L; Ramaswamy. V; Findell. KL. (2008). Strong sensitivity of late 21st
century climate to projected changes in short-lived air pollutants. J Geophys Res 113: D06102.
http://dx.doi.org/10.1029/2007JD009176
Liao. H; Chen. WT; Seinfeld. JH. (2006). Role of climate change in global predictions of future tropospheric
ozone and aerosols. J Geophys Res 111: D12304. http://dx.doi.org/10.1029/2005jd006852
Liao. H; Seinfeld. JH; Adams. PJ: Micklev. LJ. (2004b). Global radiative forcing of coupled tropospheric
ozone and aerosols in a unified general circulation model. J Geophys Res 109: D16207.
http://dx.doi.org/10.1029/2003JD004456
Logan. JA; Megretskaia. IA; Miller. AJ: Tiao. GC: Choi. D; Zhang. L; Stolarski. RS: Labow. GJ:
Hollandsworth. SM; Bodeker. GE; Claude. H; De Muer. D; Kerr. JB; Tarasick. DW: Oltmans. SJ: Johnson.
B; Schmidlin. F; Staehelin. J: Viatte. P; Uchino. O. (1999). Trends in the vertical distribution of ozone:
A comparison of two analyses of ozonesonde data. J Geophys Res 104: 26373-26399.
http://dx.doi.org/10.1029/1999JD900300
Lutter. R; Wolz. C. (1997). UV-B screening by tropospheric ozone: Implications for the national ambient air
quality standard. Environ Sci Technol 31: 142A-146A.
Madronich. S: Wagner. M; Groth. P. (2011). Influence of tropospheric ozone control on exposure to ultraviolet
radiation at the surface. Environ Sci Technol 45: 6919-6923. http://dx.doi.org/10.1021/es200701q
Marenco. A; Gouget. H; Nedelec. P: Pages. JP; Karcher. F. (1994). Evidence of a long-term increase in
tropospheric ozone from Pic du Midi data series: Consequences: Positive radiative forcing. J Geophys Res
99: 16617-16632. http://dx.doi.org/10.1029/94JD00021
Markkula. E; Salo. HM; Rikalainen. K; Jokinen. IE. (2009). Long-term UVB irradiation affects the immune
functions of carp (Cyprinus carpio) and rainbow trout (Oncorhynchus mykiss). Photochem Photobiol 85:
347-352. http://dx.doi.org/10.1111 i.1751-1097.2008.00446.x
Marquis. O: Miaud. C. (2008). Variation in UV sensitivity among common frog Rana temporaria populations
along an altitudinal gradient. Zoology (Jena) 111: 309-317. http://dx.doi.Org/10.1016/i.zool.2007.09.003
Marquis. O: Miaud. C: Lena. JP. (2008). Developmental responses to UV-B radiation in common frog Rana
temporaria embryos from along an altitudinal gradient. Population Ecology 50: 123-130.
http://dx.doi.org/10.1007/sl0144-007-0071-3
10-37
-------
Mayer. LM; Schick. LL; Hardy. KR; Estapa. ML. (2009). Photodissolution and other photochemical changes
upon irradiation of algal detritus. Limnol Oceanogr 54: 1688-1698.
Mazza. CA; Izaguirre. MM; Curiale. J: Ballare. CL. (2010). A look into the invisible: Ultraviolet-B sensitivity
in an insect (Caliothrips phaseoli) revealed through a behavioural action spectrum. Proc Biol Sci 277: 367-
373. http://dx.doi.org/10.1098/rspb.2009.1565
Meador. JA; Baldwin. AJ; Catala. P; Jeffrey. WH; Joux. F; Moss. JA; Pakulski. JD; Stevens. R; Mitchell PL.
(2009). Sunlight-induced DNA damage in marine micro-organisms collected along a latitudinal gradient
from 70 degrees N to 68 degrees S. Photochem Photobiol 85: 412-421. http://dx.doi.org/10.Ill 1/j. 1751-
1097.2008.00462.x
Menon. S: Unger. N; Koch. D; Francis. J: Garrett T; Sednev. I; Shindell. D; Streets. D. (2008). Aerosol
climate effects and air quality impacts from 1980 to 2030. Environmental Research Letters 3: 024004.
http://dx.doi.Org/10.1088/1748-9326/3/2/024004
Micklev. LJ; Jacob. DJ: Field. BD; Rind. D. (2004). Climate response to the increase in tropospheric ozone
since preindustrial times: A comparison between ozone and equivalent CO2 forcings. J Geophys Res 109:
D05106. http://dx.doi.org/10.1029/2003JD003653
Micklev. LJ: Jacob. DJ: Rind. D. (2001). Uncertainty in preindustrial abundance of tropospheric ozone:
Implications for radiative forcing calculations. J Geophys Res 106: 3389-3399.
http://dx.doi.org/10.1029/2000JD900594
Micklev. LJ: Murti. PP; Jacob. DJ: Logan. JA; Koch. DM; Rind. D. (1999). Radiative forcing from
tropospheric ozone calculated with a unified chemistry-climate model. J Geophys Res 104: 30153-30172.
http://dx.doi.org/10.1029/1999JD900439
Moss. RH; Edmonds. JA; Hibbard. KA; Manning. MR; Rose. SK; van Vuuren. DP; Carter. TR; Emori. S:
Kainuma. M; Kram. T; Meehl. GA; Mitchell. JF; Nakicenovic. N; Riahi. K; Smith. SJ: Stouffer. RJ;
Thomson. AM; Wevant. JP; Wilbanks. TJ. (2010). The next generation of scenarios for climate change
research and assessment. Nature 463: 747-756. http://dx.doi.org/10.1038/nature08823
Naik. V; Mauzerall. D: Horowitz. L; Schwarzkopf. MD; Ramaswamy. V; Oppenheimer. M. (2005). Net
radiative forcing due to changes in regional emissions of tropospheric ozone precursors. J Geophys Res
110: D24306. http://dx.doi.org/10.1029/2005JD005908
Naja. M; Akimoto. H. (2004). Contribution of regional pollution and long-range transport to the Asia-Pacific
region: Analysis of long-term ozonesonde data over Japan. J Geophys Res 109: D21306.
http://dx.doi.org/10.1029/2004JD004687
NRC (National Research Council). (2005). Radiative forcing of climate change: Expanding the concept and
addressing uncertainties. Washington, DC: The National Academies Press.
http://books.nap.edu/openbook.php7record id=11175&page=Rl
Obara. Y; Koshitaka. H; Arikawa. K. (2008). Better mate in the shade: Enhancement of male mating behaviour
in the cabbage butterfly, Pieris rapae crucivora, in a UV-rich environment. J Exp Biol 211: 3698-3702.
http://dx.doi.org/10.1242/ieb.021980
Oltmans. SJ: Lefohn. AS: Harris. JM; Galbally. I; Scheel. HE; Bodeker. G: Brunke. E; Claude. H; Tarasick. D;
Johnson. BJ; Simmonds. P; Shadwick. D; Anlauf. K; Havden. K; Schmidlin. F; Fujimoto. T; Akagi. K;
Meyer. C: Nichol. S: Davies. J: Redondas. A; Cuevas. E. (2006). Long-term changes in tropospheric ozone.
Atmos Environ 40: 3156-3173. http://dx.doi.Org/10.1016/i.atmosenv.2006.01.029
Ordonez. C: Brunner. D; Staehelin. J: Hadjinicolaou. P: Pyle. JA; Jonas. M; Wernli. H; Prevot. ASH. (2007).
Strong influence of lowermost stratospheric ozone on lower tropospheric background ozone changes over
Europe. Geophys Res Lett 34: L07805. http://dx.doi.org/10.1029/2006GL029113
Oromi. N; Marquis. O: Miaud. C: Sanuy. D. (2008). Influence of ambient ultraviolet radiation on Bufo
calamita egg development in a semiarid zone (Catalonia, Spain). J Environ Biol 29: 135-137.
10-38
-------
Pavelin. EG: Johnson. CE; Rughooputh. S: Toumi. R. (1999). Evaluation of pre-industrial surface ozone
measurements made using Schonbein's method. Atmos Environ 33: 919-929.
http://dx.doi.org/10.1016/S1352-2310(98)00257-X
Phoenix. GK; Gwynn-Jones. D; Lee. JA; Callaghan. TV. (2000). The impacts of UV-B radiation on the
regeneration of a sub-arctic heath community. Plant Ecol 146: 67-75.
http://dx.doi.Org/10.1023/A:1009839506658
Pickett. JE; Gibson. DA; Gardner. MM. (2008). Effects of irradiation conditions on the weathering of
engineering thermoplastics. Polym Degrad Stabil 93: 1597-1606.
http://dx.doi.0rg/10.1016/i.polymdegradstab.2008.02.009
Price. C: Fund. D. (1994). Possible implications of global climate change on global lightning distributions and
frequencies. J Geophys Res 99: 10823-10831. http://dx.doi.org/10.1029/94JD00019
Pyle. JA; Warwick. N; Yang. X; Young. PJ; Zeng. G. (2007). Climate/chemistry feedbacks and biogenic
emissions. Philos Transact A Math Phys Eng Sci 365: 1727-1740. http://dx.doi.org/10.1098/rsta.2007.2041
RCP (Representative Concentration Pathways). (2009). RCP Database (version 2.0). Available online at
http://iiasa.ac.at/web-apps/tnt/RcpDb/dsd?Action=htmlpage&page=about (accessed January 28, 2011).
Riggsbee. JA; Orr. CH; Leech. DM; Doyle. MW: Wetzel. RG. (2008). Suspended sediments in river
ecosystems: Photochemical sources of dissolved organic carbon, dissolved organic nitrogen, and adsorptive
removal of dissolved iron. J Geophys Res 113: G03019. http://dx.doi.org/10.1029/2007jg000654
Rind. D: Healy. R; Parkinson. C: Martinson. D. (1995). The role of sea ice in 2CO2 climate model sensitivity.
Part I: The total influence of sea ice thickness and extent. J Clim 8: 449-463.
http://dx.doi.org/10.1175/1520-0442(1995)008<0449:TROSII>2.0.CO:2
Romansic. JM; Waggener. AA; Bancroft. BA; Blaustein. AR. (2009). Influence of ultraviolet-B radiation on
growth, prevalence of deformities, and susceptibility to predation in Cascades frog (Rana cascadae) larvae.
Hydrobiologia 624: 219-233. http://dx.doi.org/10.1007/sl0750-009-9703-2
Sanderson. MG: Jones. CD; Collins. WJ: Johnson. CE; Derwent. RG. (2003). Effect of climate change on
isoprene emissions and surface ozone levels. Geophys Res Lett 30: 1936.
http://dx.doi.org/10.1029/2003GL017642
Schindler. DW: Curtis. PJ: Parker. BR; Stainton. MP (1996). Consequences of climate warming and lake
acidification for UV-B penetration in North American boreal lakes. Nature 379: 705-708.
http://dx.doi.org/10.1038/379705aO
Semerdjieva. SI; Phoenix. GK; Hares. D: Gwynn-Jones. D: Callaghan. TV; Sheffield. E. (2003). Surface
morphology, leaf and cuticle thickness of four dwarf shrubs from a sub-Arctic heath following long-term
exposure to enhanced levels of UV-B. Physiol Plant 117: 289-294. http://dx.doi.org/10.1034/U399-
3054.2003.00006.x
Shindell. D: Faluvegi. G. (2009). Climate response to regional radiative forcing during the twentieth century.
Nat Geosci 2: 294-300. http://dx.doi.org/10.1038/ngeo473
Shindell. D: Faluvegi. G: Lacis. A; Hansen. J: Ruedv. R: Aguilar. E. (2006). Role of tropospheric ozone
increases in 20th-century climate change. J Geophys Res 111: D08302.
http://dx.doi.org/10.1029/2005JD006348
Shindell. DT; Faluvegi. G. (2002). An exploration of ozone changes and their radiative forcing prior to the
chlorofluorocarbon era. Atmos Chem Phys Discuss 2: 363-374. http://dx.doi.org/10.5194/acp-2-363-2002
Shindell. DT; Faluvegi. G: Bauer. SE; Koch. DM; Unger. N; Menon. S: Miller. RL; Schmidt. GA; Streets. DG.
(2007). Climate response to projected changes in short-lived species under an A1B scenario from 2000-
2050 in the GISS climate model. J Geophys Res 112: D20103. http://dx.doi.org/10.1029/2007jd008753
Shindell. DT; Faluvegi. G: Bell. N. (2003). Preindustrial-to-present-day radiative forcing by tropospheric
ozone from improved simulations with the GISS chemistry-climate GCM. Atmos Chem Phys 3: 1675-
1702. http://dx.doi.org/10.5194/acp-3-1675-2003
10-39
-------
Shindell. DT; Lew H. II: Schwarzkopf. MD; Horowitz. LW: Lamarque. JF; Faluvegi. G. (2008). Multimodel
projections of climate change from short-lived emissions due to human activities. J Geophys Res 113:
D11109. http://dx.doi.org/10.1029/2007JD009152
Simmonds. PG: Dement. RG: Manning. AL; Spain. G. (2004). Significant growth in surface ozone at Mace
Head, Ireland, 1987-2003. Atmos Environ 38: 4769-4778.
http://dx.doi.0rg/10.1016/i.atmosenv.2004.04.036
Sinha. A; Toumi. R. (1997). Tropospheric ozone, lightning, and climate change. J Geophys Res 102: 10667-
10672. http://dx.doi.org/10.1029/96JD03710
Sitch. S: Cox. PM; Collins. WJ: Huntingford. C. (2007). Indirect radiative forcing of climate change through
ozone effects on the land-carbon sink. Nature 448: 791-794. http://dx.doi.org/10.1038/nature06059
Slinev. DH; Wengraitis. S. (2006). Is a differentiated advice by season and region necessary? Prog Biophys
Mol Biol 92: 150-160. http://dx.doi.Org/10.1016/i.pbiomolbio.2006.02.007
Snell. KRS: Kokubun. T; Griffiths. H; Convey. P; Hodgson. DA; Newsham. KK. (2009). Quantifying the
metabolic cost to an Antarctic liverwort of responding to an abrupt increase in UVB radiation exposure.
Global Change Biol 15: 2563-2573. http://dx.doi.Org/10.llll/i.1365-2486.2009.01929.x
Soden. BJ; Held. IM. (2006). An assessment of climate feedbacks in coupled ocean-atmosphere models. J Clim
19:3354-3360.
Staehelin. J: Thudium. J: Buehler. R; Volz-Thomas. A; Graber. W. (1994). Trends in surface ozone
concentrations at Arosa (Switzerland). Atmos Environ 28: 75-87. http://dx.doi.org/10.1016/1352-
2310(94)90024-8
Stevenson. DS. (2004). Radiative forcing from aircraft NO emissions: Mechanisms and seasonal dependence. J
Geophys Res 109: D17307. http://dx.doi.org/10.1029/2004JD004759
Stevenson. DS: Dentener. FJ; Schultz. MG: Ellingsen. K; Van Noije. TPC: Wild. O: Zeng. G: Amann. M;
Atherton. CS: Bell. N; Bergmann. DJ; Bey. I; Butler. T; Cofala. J: Collins. WJ: Derwent. RG: Dohertv.
RM; Drevet. J: Eskes. HJ; Fiore. AM; Gauss. M; Hauglustaine. DA; Horowitz. LW: Isaksen. ISA; Krol.
MC: Lamarque. JF; Lawrence. MG: Montanaro. V; Muller. JF; Pitari. G: Prather. MJ; Pyle. JA; Rast. S:
Rodriguez. JM; Sanderson. MG: Savage. NH; Shindell. DT; Strahan. SE; Sudo. K; Szopa. S. (2006).
Multimodel ensemble simulations of present-day and near-future tropospheric ozone. J Geophys Res 111:
D08301. http://dx.doi.org/10.1029/2005JD006338
Streets. D; Bond. T; Lee. T; Jang. C. (2004). On the future of carbonaceous aerosol emissions. J Geophys Res
109: D24212. http://dx.doi.org/10.1029/2004JD004902
Tanimoto. H. (2009). Increase in springtime tropospheric ozone at a mountainous site in Japan for the period
1998-2006. Atmos Environ 43: 1358-1363. http://dx.doi.Org/10.1016/i.atmosenv.2008.12.006
Tarasick. DW: Fioletov. VE; Wardle. PI; Kerr. JB; Davies. J. (2005). Changes in the vertical distribution of
ozone over Canada from ozonesondes: 19802001. J Geophys Res 110: D02304.
http://dx.doi.org/10.1029/2004JD004643
Thompson. AM. (1992). The oxidizing capacity of the Earth's atmosphere: Probable past and future changes.
Science 256: 1157-1165. http://dx.doi.org/10.1126/science.256.5060.1157
Thompson. AM; Chappellaz. JA; Fung. IY; Kucsera. TL. (1993). The atmospheric CH4 increase since the last
glacial maximum (2) Interactions with oxidants. Tellus B Chem Phys Meteorol 45: 242-257.
http://dx.doi.Org/10.1034/i.1600-0889.1993.t01-2-00003.x
Thompson. AM; Hudson. RD. (1999). Tropical tropospheric ozone (TTO) maps from Nimbus 7 and Earth
Probe TOMS by the modified-residual method: Evaluation with sondes, ENSO signals, and trends from
Atlantic regional time series. J Geophys Res 104: 26961-26975. http://dx.doi.org/10.1029/1999JD900470
U.S. EPA (U.S. Environmental Protection Agency). (2003). National ambient air quality standards for ozone:
Final response to remand. Fed Reg 68: 614-645.
10-40
-------
U.S. EPA (U.S. Environmental Protection Agency). (2006b). Air quality criteria for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/R-05/004AF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=149923
UNEP (United Nations Environment Programme). (2009). Environmental effects of ozone depletion and its
interactions with climate change. Nairobi, Kenya. http://ozone.unep.org/Assessment Panels/EEAP/EEAP-
Progress-report-2009.pdf
Unger. N. (2006). Cross influences of ozone and sulfate precursor emissions changes on air quality and
climate. PNAS 103: 4377-4380. http://dx.doi.org/10.1073/pnas.0508769103
Unger. N; Bond. TC: Wang. JS: Koch. DM; Menon. S: Shindell. DT; Bauer. S. (2010). Attribution of climate
forcing to economic sectors. PNAS 107: 3382-3387. http://dx.doi.org/10.1073/pnas.0906548107
Unger. N; Shindell. DT; Koch. DM; Streets. DG. (2008). Air pollution radiative forcing from specific
emissions sectors at 2030. J Geophys Res 113: D02306. http://dx.doi.org/10.1029/2007JD008683
Van Aardenne. JA; Dentener. FJ; Olivier. JGJ: Klein Goldewijk. CGM; Lelieveld. J. (2001). A 11 resolution
data set of historical anthropogenic trace gas emissions for the period 1890-1990. Global Biogeochem
Cycles 15: 909-928. http://dx.doi.org/10.1029/2000GB001265
Volz. A; Kiev. D. (1988). Evaluation of the Montsouris series of ozone measurements made in the nineteenth
century. Nature 332: 240-242. http://dx.doi.org/10.1038/332240aO
Wahl. M. (2008). Ecological modulation of environmental stress: Interactions between ultraviolet radiation,
epibiotic snail embryos, plants and herbivores. J Anim Ecol 77: 549-557. http://dx.doi.org/10.1111/i. 1365-
2656.2007.01352.x
Wang. WC: Pinto. JP; Yung. YL. (1980). Climatic effects due to halogenated compounds in the earth's
atmosphere. J Atmos Sci 37: 333-338. http://dx.doi.org/10.1175/1520-
0469(1980)037<0333:CEDTHC>2.0.CO:2
West. JJ: Fiore. AM. (2005). Management of tropospheric ozone by reducing methane emissions [Review].
Environ Sci Technol 39: 4685-4691.
West. JJ: Fiore. AM; Horowitz. LW: Mauzerall. PL. (2006). Global health benefits of mitigating ozone
pollution with methane emission controls. PNAS 103: 3988-3993.
http://dx.doi.org/10.1073/pnas.0600201103
West. JJ: Fiore. AM; Naik. V; Horowitz. LW: Schwarzkopf. MD; Mauzerall. PL. (2007). Ozone air quality and
radiative forcing consequences of changes in ozone precursor emissions. Geophys Res Lett 34: L06806.
http://dx.doi.org/10.1029/2006GL029173
Wild. O: Prather. MJ; Akimoto. H. (2001). Indirect long-term global radiative cooling from NOX emissions.
Geophys Res Lett 28: 1719-1722. http://dx.doi.org/10.1029/2000GL012573
Worden. HM; Bowman. KW: Worden. JR; Eldering. A; Beer. R. (2008). Satellite measurements of the clear-
sky greenhouse effect from tropospheric ozone. Nat Geosci 1: 305-308. http://dx.doi.org/10.1038/ngeol82
Wu. S: Micklev. LJ; Jacob. PJ: Rind. P; Streets. PG. (2008a). Effects of 2000-2050 changes in climate and
emissions on global tropospheric ozone and the policy-relevant background surface ozone in the United
States. J Geophys Res 113: P18312. http://dx.doi.org/10.1029/2007JP009639
Wu. S: Micklev. LJ: Leibensperger. EM; Jacob. PJ: Rind. P; Streets. PG. (2008b). Effects of 2000-2050
global change on ozone air quality in the United States. J Geophys Res 113: P06302.
http://dx.doi.org/10.1029/2007JP008917
Yang. X; Cox. RA; Warwick. NJ; Pyle. JA; Carver. GP; O'connor. FM; Savage. NH. (2005c). Tropospheric
bromine chemistry and its impacts on ozone: Amodel study. J Geophys Res 110: P23311.
http://dx.doi.org/10.1029/2005JP006244
Zeng. G: Pyle. JA. (2003). Changes in tropospheric ozone between 2000 and 2100 modeled in a chemistry-
climate model. Geophys Res Lett 30: 1392. http://dx.doi.org/10.1029/2002GL016708
10-41
-------
Zepp. RG: Erickson. DJ; Paul. ND; Sulzberger. B. (2007). Interactive effects of solar UV radiation and climate
change on biogeochemical cycling. In The Environmental Effects of Ozone Depletion and its Interactions
with Climate Change: 2006 Assessment (pp. 135-164). Nairobi, Kenya: United Nations Environment
Programme.
Zepp. RG: Shank. GC: Stabenau. E; Patterson. KW: Cvterski. M; Fisher. W: Bartels. E; Anderson. SL. (2008).
Spatial and temporal variability of solar ultraviolet exposure of coral assemblages in the Florida Keys:
Importance of colored dissolved organic matter. Limnol Oceanogr 53: 1909-1922.
Zerefos. CS: Kourtidis. KA; Melas. D; Balis. D; Zanis. P; Katsaros. L; Mantis. HT; Repapis. C: Isaksen. I;
Sundet. J: Herman. J; Bhartia. PK; Calpini. B. (2002). Photochemical activity and solar ultraviolet radiation
(PAUR) modulation factors: An overview of the project. J Geophys Res 107: 8134.
http://dx.doi.org/10.1029/2000JD000134
Ziemke. JR; Chandra. S: Bhartia. PK. (2005). A 25-year data record of atmospheric ozone in the Pacific from
Total Ozone Mapping Spectrometer (TOMS) cloud slicing: Implications for ozone trends in the
stratosphere and troposphere. J Geophys Res 110: Dl5105. http://dx.doi.org/10.1029/2004JD005687
10-42
------- |